I put the radiator in the wrong end of a Rover P5

You alright there? Me too. I've been busy.

Previously, I wrote:

and anyway nobody said that the radiator must be at the front of the car.

:)

My thoughts are like wild animals. Sometimes they're exciting, sometimes they're cute, and it's best not to leave two of them in a room with only each other to feed on.

For example, let's look at the boot of a tatty Rover P5.

Normal people would say something like "look at the size of that! There's lots of room for suitcases and shopping or whatever the fuck normal people put in their cars". I, on the other hand, thought "look how much room there is for a radiator".

So, let's talk about that. Basically every car with the engine in the front have the radiator in the front of the car too. That means the coolant lines are short and simple, and means that the boot can be used for suitcases and so on.

Alternatively you can put it in the back of the car. Here is a rear-mounted radiator in Matthew Holder's E36 Compact.

Putting it in the back of the car, i.e. in the boot, has some overwhelming advantages which make it common in torture-test applications like drifting. The primary benefit is moving the primary source of heat (your engine) away from the primary means of removing heat (your radiator). This helps mitigate heat soak, which I'll crudely summarise as "everything getting hot and keeping everything else hot".

There is a smaller benefit that you have much longer coolant lines, and those lines contain coolant, and that adds to the total volume of coolant in the system. A pair of three-metre long coolant lines an inch in diameter will increase coolant capacity by about 3 litres, or roughly a quarter of the capacity of the entire coolant system of the car the engine came out of.

There are disadvantages to this. You lose luggage space; but if I want to move anything I would rather use a practical car. The coolant lines are longer and more complicated. And invariably this means the radiator is no longer the highest point in the cooling system. Some people believe that makes it more difficult to bleed, but this is a solved problem.

Anyway that's how I ended up cutting out the boot floor of a Rover P5. But before I could cut any of it out, I wanted to remove whatever this stuff was from it.

This needed to be removed if I was to weld anything to the remainder of the boot floor once I cut it out, and because I was planning to progressively cut out the floor and reinforce it I would have to remove it before doing any cutting. This stuff also made it hard to mark any cut lines on the floor.

It seemed like bitumen; whatever it is my health is likely not better for having been in its vicinity. It can't be ground off with my usual tools like various drill attachments or angle grinder things, because they just heat it up and spread it around, and give you a differently-shaped mess to remove. You'd never get a surface clean enough to weld to. And manual means like wire brushes and scrapers just seemed to bounce off it.

What helped here was freeze spray, which is apparently used for diagnosing electronics. This makes it brittle for just long enough to remove the bulk of it with a hammer and chisel and by the way this took forever.

Originally, as said, my plan was to temporarily brace the floor with angle iron joining the edge which I was cutting out, and then replace it progressively with 25x2.5mm steel box section. You can see some of that temporary bracing on the front edge below.

It turned out to not be necessary; my assumption that the boot floor was critically structural was invalidated when I started cutting it out. It was so floppy that I could flex it between two fingers. Still, I decided to continue with the reinforcement. It adds some rear-end rigidity (especially with two-inch gussets), probably doesn't add any weight, gives me somewhere to build the radiator mounts from, and looks better than an unfinished hole.

But while welding that in, I became increasingly aware that I wasn't welding this to very much. That was because the area where the battery box used to be (which is not where the battery is going to be) was, quite predictably, entirely rotten.

That started a side quest...

...of replacing about a foot and a half of steel, and there went an entire day.

The other side looked pretty grim, but after cleaning up most of it was pretty solid.

It only needed most of the upward return (to which I was welding the reinforcing frame) cutting out and replacing. (To Internet welders: I know. It'll hold.)

Anyway, the point of this wasn't just to make a big gaping hole in the boot of my car. It was to make a big gaping hole to get some kind of under-body airflow to a radiator, and after a quick chat with Chris from Pro Alloy (who previously remade the intercooler for my 323 GTX) the radiator I chose was this:

It's Pro Alloy's radiator for the Ford Sierra Cosworth. Any radiator I bought would only be a starting point (I'll want to modify the inlets and outlets); this one was priced well enough and I know it'll be up to the job.

What I could do here is use the original mounts for the Cosworth radiator. There are two of them, and they look like these ones from Burton.

In fact I did buy two of these, and they made it into the final product, because the mounting points existed on the radiator and I might as well use them. But I do not trust them to support the radiator alone in this application. This is partly because they were never designed to support the radiator at a 45 degree angle, which is how it will be mounted in the P5.

Instead, I needed to make a mounting frame for the radiator, and let's talk about dimple dies.

Dimple dies are cool! I don't know why I waited this long to get some! You take a boring piece of steel, drill a hole in it...

...stick a dimple die through it, apply about 5-10 tons of force in a hydraulic press, and...

...you are now RACE CAR FABRICATOR. It looks cool, and because a hole with no metal is lighter than metal it saves weight, without compromising strength.

And then if you're me you absolutely go to town on it.

So the frame is made of some 50mmx50mmx2mm steel angle on the sides, with some 20mm box section steel on the top and bottom joining them.

Attaching to that are various threaded mounting points for the radiator itself (using the top and bottom flanges on the radiator) and for a fan shroud. These were made from offcuts of the same massive thick angle iron I used to brace my axle housing when I was welding stuff to it.

And on each corner is a square section of the same 50mm steel angle with a slotted hole for some generic rubber anti-vibration mounting bushes.

This is critical! It doesn't just isolate vibration; it stops the assembly becoming a stressed member when the bodyshell flexes. If it wasn't flexibly mounted I'd have to make the frame out of much heavier material to stop the radiator becoming stressed and cracking.

Scattered around are some of these cute little weld-on tabs for cable-tying a wiring harness to. I think I know how my wiring will be routed, so I committed to it.

This radiator-holding frame mounts to the body via these brackets which are to be welded on to the frame in my boot. It's made of the same 50mm angle iron used earlier, and the remnants of the 1" box section I used to line the hole I made in my boot floor. The box section is rotated through 45 degrees to impart the same tilt on the radiator when it is mounted.

Except that's boring, so let's have some more dimple dies, because race car.

Better!

With those welded into the boot I could do the first all-up test.

Everything landed more or less where I thought it would! Especially the outlet on the bottom; this ended up just below the boot floor, which is exactly where I wanted them. The mounting rubber closest to the camera was a little over-stressed in the shear direction; I remedied that by grinding the slot wider and welding in a little shim.

Those who know Sierra Cosworth radiators (this is pretty niche) will notice that the radiator is actually mounted upside down. On a Cosworth, the fan switch (which might get replaced with a temperature sensor) is installed on the inlet (hot) side of the radiator. On mine, the radiator is flipped so that the fan switch is mounted on the outlet side of the radiator. This is because I, one guy on the Internet, disagree with Cosworth, engineering geniuses. I don't care what temperature my water is going in to the radiator. It might come out a different temperature to what it went in, e.g. if the radiator is doing what a radiator does, and that temperature should dictate whether the fans need to be switched on or nmot.

At this point the mounting frame just looks like random bits of steel crudely stuck together by some idiot with a MIG, because it is. But with all the welds tidied up and a coat of Jenolite satin black it looked way better than random bits of steel stuck together by an idiot with a MIG has any right to look.


This is all to be cooled by twin 10 inch fans from American Cooling Systems, each of which is allegedly rated for 2187 CFM. That's a lot of fan, but to make it more efficient it is best to have them mounted via a fan shroud so that they pull air through the entire radiator. I mentioned in passing earlier that I was going to fit one. Unfortunately, I couldn't find anyone who makes a fan shroud for Rover P5s with rear-mounted Sierra Cosworth radiators. Strange! You'd have thought that'd be an off-the-shelf part!

That meant I had to make one, and because I was going to make one I thought I'd make one a little bit special.

That's right, actual carbon fibre.

I didn't take any step-by-step pics of this, but this should be fairly self-explanatory: 20mm carbon fibre box section for the standoff, 2mm carbon fibre sheet for the face, and 25x2mm carbon fibre angle for the mounting points, glued together with VM100 adhesive. The fans bolt through with those weird glued-in threaded stud thingies. You can get all of this from Easy Composites, who happen to have an outstanding YouTube channel.

Cured carbon fibre is nicer than I expected to work with! It cuts easily with a normal angle grinder cutting disc or a Dremel-like tool. Easy Composites have a video on the subject. I'm not scared of it anymore, so there might be more carbon fibre parts in my future.

This is crude, and my first attempt at making anything with carbon fibre, but it looks absolutely sick. And nobody will ever see it because it faces the floor! But I know it's there, and it means I get to say "carbon fibre fan shroud".


And that, plus some foam and rubber stripping, is a radiator mounted in the wrong end of a Rover P5.

There is still much to do, because the rest is actually plumbing in the entire cooling system. The radiator needs some modification to the inlets and outlets accommodate a 45 degree mounting angle, an expansion tank needs mounting somewhere, and of course there are no coolant lines.

And in fact I'm not 100% sure there's going to be enough air flow available here. There should be enough underneath to allow cooling at speed, and once I ventilate the boot lid (not at all sorry to anyone who hates seeing perfectly good P5 panels getting chopped up) I should have enough air around for the fans to pull from. That is guesswork, which is not the same as knowing. Without putting this in a wind tunnel, I won't know how well this will work until I actually use it, and it does or does not overheat.

But that's a problem for future Lewis. If I look at everything that I need to do to this car I can be overwhelmed; someone who doesn't really know what the fuck he is doing re-engineering an entire car. But if I break it down into lots of small, standalone projects like this I find it much more manageable, and doing them in whatever order I find most fun on a given day keeps the motivation high.


I know what you're thinking. It's something like "Lewis, this is all totally unnecessary and overkill unless you were planning on, e.g., bolting a massive turbo onto it and getting stupid horsepower numbers which would cause you to actually need all this cooling".

I know.

:)


Part numbers from this post:

  • Freeze spray:: Motip 090409
  • Radiator: Pro Alloy RADCOSS
  • Boot floor corner repair section: J R Wadhams P5BDW37
  • Satin black paint: Jenolite 89036
  • Red primer: Jenolite 89593
  • Carbon fibre: Easy Composites CFS-RI-2-0475, CFBOX-20-17-2, CFANG-25-2-1
  • Cable tie mount thingies: Pacific Customs AC750795
  • Steel by Thomas B Bonnett, KI Metals and some random eBay seller

A new axle for the Rover P5, part 2: the housing

I know what you're thinking. It's something like "Lewis, a Rover P5's prehistoric rear axle will explode if you put 350 horsepower through it". And you're right!

There are lots of ways I could build a rear end for the P5 capable of handling LS power. Probably the sensible thing to do would be to chuck a BMW E60 subframe under the back and call it solved. That'd be cheap, parts would be available long after I wrap the car around a tree, and the fabrication doesn't scare me too much. But while I'm hardly shooting for originality here, fully-independent rear suspension crosses a line for me where the car wouldn't really be a P5 anymore.

Instead, I wanted a live axle like the one that came out of of the P5, and the inevitable conclusion of all discussion of live axles is the Ford 9-inch.

I don't have a photo of a stock Ford 9 inch axle to use here; take this and fill in the gaps with your imagination.

The Ford 9-inch is a design that goes back to the late 1950s, was phased out in the mid 1980s and was fitted to millions of cars and trucks in the United States in those 30-ish years. It is legendarily strong; a completely stock axle is safe up to about 400 horsepower, and axles built with the right aftermarket parts will survive "you'd have to be utterly stupid to use this on the road" horsepower. And the aftermarket is amazing! It's such a sound, well-tested and well-loved design on the other side of the Atlantic that you could build a complete brand-new "Ford" 9-inch axle without a single Ford component, if you were mad.

So here is my starting point.

This is a Ford 9-inch axle housing that does not contain a single Ford component. Instead of any Ford components, it is a Strange Engineering heavy-duty housing, welded to 3" CDS axle tubes, with Strange Engineering "late Big Ford" housing ends. It's a work of art, the work to assemble it into the housing is perfect, and it is good for more power than even an unreasonable person would try to squeeze out of an LS1.

I wish I could boast that this housing was my work, but it isn't. I can stick two pieces of metal together with a welder such that they don't break, but I can't do that while holding an axle housing and axle tube dead straight on all axes. And even if I could the welds wouldn't look slightly as nice as this.

It was instead built for me by Hauser Racing, to the correct width to fit under the Rover P5. I love Hauser Racing! They're not a mere importer of parts, though they can in fact get all the parts from the States that I could possibly want. They know the things they sell inside and out because they build hardcore drag cars, and by the way are literally a drag racing team too whose founder is a bit of a legend in that world. When I call them, I immediately have someone on the other end who can answer my technical questions. Most of those questions were basic stuff as I navigated the new-to-me world of overkill rear axles, so they're also very nice for putting up with me.


So I had a beautifully-welded axle housing, and I had two things to weld on to it. Both of these were spring perches that I made in part one.

The first job was to work out what their position should be, and the easiest measurement to take is from the outmost edges of the spring perches. I measured this on two different Rover axle housings: my original one after I took it off the car, and the one that I got from David Green. These actually measured slightly differently. My original measured 1140mm, and the one from David Green measured as 1145 mm. A curiosity was that the one from David Green had one spring perch 5mm closer to the ends of the axle tubing than the other, for unclear reasons (it does not look like it was ever chopped off and welded back on again, ruling out the obvious explanation).

To settle it, I took a couple of measurements from my P5's shell; these suggested that 1145 was probably correct for my car. It might not be for yours. The easiest way to do this is to measure the distance between the outside edges of your chassis rail immediately above your axle, subtract the width of one chassis rail (this gives you the distance between the centres of your chassis rails), then add the width of one of your leaf spring perches (this is 80mm for mine; it's probably 79mm on the originals).

And I know, it seems absurd to chase millimetres that the car wasn't made with in the first place (see also the 5mm difference between two different unmodified axles!). The slop in the suspension of an old leaf-sprung car seems like it'll outweigh all of this. But two things:

  1. Polybushes and a radically different rear leaf spring bush replacement (this last to be discussed in a future post) will hopefully engineer much of the forgiveness out of the rear end, reducing the tolerances I have to work with;
  2. If you don't have the resources of a real fabricator (this is mostly me and an angle grinder and a 40 year old bandsaw and a free MIG welder) you are going to end up with some things out of tolerance no matter how hard you try. So chase millimetres while you can and don't use up the tolerances when you don't need to!

My axle housing is 1384 to its very ends. So that means the outside of my spring perches needed to be 119mm from the Ford housing's axle ends. That's easy to do with a set square. To make life easy for myself, I temporarily held the perches in place with magnets, validated the measurements then marked along the perches' sides with a paint pen.

That means that I can eyeball the perches' proper position when I put them in place to be welded. That is useful because there's one more thing to be set up, which is their angle. These must not be dead flat on the bottom edge! My P5, the original P5, and just about every rear-wheel-drive car has a 3 degree downward tilt (towards the rear) to the entire drivetrain. To eliminate vibration, the two faces of the propshaft must be parallel. That means that we also must point the input shaft of the axle slightly upwards, by the same 3 degrees as the drivetrain angle. You can do this with shims after the fact, but I found out (thanks Hauser Racing for fielding my dumbass calls!) that it is much better to do this by positioning your leaf spring perches appropriately.

This is fiddly to set up; a massive vice to hold your axle still and a calibrated digital level are your friends. And this is why I marked the left-to-right location of the perches first. It meant I could eyeball that dimension any time I needed it and concentrate on getting the rotation right.

Next up was welding them on. The enemy here was heat distortion. Welding generates a lot of heat, and heat causes metal to expand, contract, and move. This often doesn't matter for things that aren't axle housings; you can always bash it back into shape or just start again. An axle housing doesn't have much tolerance for misalignment, I don't have the means of beating it back into shape accurately, and "start again" would make me four figures poorer.

To solve this problem, here is a piece of angle iron sitting atop a piece of tube.

This is how angle iron will always want to sit on a cylinder. You can exploit this property should you ever need to draw a dead straight line on a piece of tubing; just put a piece of angle iron on top of it and draw along its edge. You're welcome! My theory was that you could also exploit this property to keep a piece of tube dead straight while welding it; if you clamp a massive section of angle iron really tight to a piece of tube it will resist distortion on two axes.

If you clamp a massive piece of angle iron to one side of the tube, and an extra piece to the other side, and you only weld a little at a time and let it cool fully, you should have pretty good odds of avoiding heat distortion.

That's the theory I probably didn't fully think through! But even now it seems sound to me. We'll find out when my driveshafts do or do not fit properly.

G-clamps do not naturally want to sit on the apex of a piece of angle iron. I could have welded little flat pieces onto the angle iron. But then I might distort the angle iron while welding it, and it would no longer be as completely straight as I needed it to be for this purpose. Instead, I got some scraps out of the "might be useful some day" bin and made these little adapter pieces.

It doesn't matter if these bits distort while I welded them, because they aren't references. They just need to sit on the bigger piece of angle iron tightly enough that the bigger piece wouldn't move, and to give a mostly-flat edge for the G-clamp to clamp to. I only had one G-clamp that was actually big enough to hold one of these adapters on each side. Instead, I MIGed some blobs of metal onto one of the faces of two of my worse, cheaper clamps and that stopped them from sliding off the apex when I tightened them down.

The first perch took the best part of a day, but that included making up those little adapter pieces. It took about 20 minutes for each weld and the surrounding tubing to fully cool to ambient temperature. It was slow going, but slow going was much more fun than writing off a very fancy axle housing. But when I was done, with all the welds tidied up...

...it looked pretty damn sweet. I don't normally bother smoothing out welds like this; I don't much care to pretend that I'm a better welder than I really am. But in this case I had to smooth out the welds at all the corners, otherwise the U-bolts won't clear. And if I'm doing that some Lewis-logic said I might as well get completely carried away and do all of the welds to make them look pretty.

Anyway, do all that twice...

...and you have a Ford 9-inch axle housing with two home-made leaf spring perches for a Rover P5. Cool!

It doesn't have a diff or driveshafts yet. This is already in the works (thanks Hauser!). And it doesn't have any fixtures for routing the brake lines. But to be able to make those I would need some brakes, because otherwise I would just be guessing where those fixtures needed to be. And that's why I called Hauser Racing again, and got talked into buying a box...

...containing a brake kit for the Ford 9" axle, with 11-inch vented discs and Wilwood four-pot calipers. But I'll tell you about that in part 3.

See you next time.

:)

Remaking Rover P5 rear leaf spring perches

These are Rover P5 leaf spring perches, which aren't actually from my Rover P5.

These weld to the axle and have the leaf spring clamped to the bottom of them. I needed some of these because I will be running standard rear suspension on my LS-engined P5, but I will be using a completely different axle housing, about which there will be more very soon. But what I didn't want to do, was chop up my old axle for these parts.

Doing that would be a shame, because it would have written off the axle housing. And this is a complete, original Mark 1A export CKD rear axle in excellent condition. I would rather an entire axle be saved for someone that needs it.

Instead, I rented a van (in which I learned that I don't like modern vehicles, but do like their fuel economy), took a trip to everyone's favourite supplier of unobtainable P5 parts...

...and came home with an axle housing. And also a Pitman arm, and a door which is in much worse condition than any the one I have. If you wonder why I would specifically seek out a door that is not in good condition...well there are Reasons, and at least one of those will be made clear some time in the next year.

Anyway, while I didn't want to write off an entire axle, I didn't mind cutting up an axle housing that had been kicking around for a while.

Cutting the perches off was easy, because I didn't care about preserving the axle housing itself. It also smelled really good, because this was kicking around outdoors, and so was full of leaves rather than grease. I briefly had the world's largest incense burner in the workshop!

Closer inspection showed that these needed a little repairing; there was a hole on the outside where the metal had thinned out.

I plated up the inside and filled in with the MIG from the outside, and that was a lot better.

There was a piece on the inside where this had been repaired previously, and badly.

I tried touching up that repair, and then had a side quest upgrading my welder. And after the welder returned I got these perches sandblasted to see what I was actually working with.

That exposed a flaw in this particular plan, which is that they were Fucked.

The plate had pinholes, it was dangerously thin in much of the rest of it, the shock absorber mounting rod had thinned down by 4mm in places, and the bad repair was even worse than it looked when I first saw it. If it had only one of these problems it might be salvageable, if I was inclined to do it, and I'd be more inclined to do it if it wasn't impossible to get a MIG nozzle down there at any angle but vertical.

I think it's important to point out that this is not the fault of the man I got this housing from. David is well-respected in the P5 community for a good reason, and I think he's thoroughly honest and nice. This one's on me; I agreed to take it unseen and there was nothing obviously wrong with it at a brief inspection. It was up to me to figure out whether this would be good enough for what I wanted to do with it, and I didn't.

So like, plan B.

Like, Plan B

I'd have liked some other Plan B than the one I ended up with, such as buying replacements, but you can't. They don't exist for the P5, and there's nothing else I could use either. Even if there were other vehicles with 3 inch wide leaf springs for 3 inch axle tubes, I doubt any would have the shock absorber mount built in and in the correct place. Instead, I had to make them. Because how hard can it be!

It starts with some 80x80mm box section.

That seems like a curiously metric dimension for a car that uses Imperial measurements, because it is. But I measured the original perches at a hair over 79mm wide. Because of how tight the original leaf spring U-bolts were against the perches, I might have to slot the holes in the bottom plate by half a millimeter; we'll see when I come to fit all this together. I will not bother to recreate the rib in the middle as found on the original perches. It's an unnecessary complication; it isn't needed with steel this thick.

I cut two pieces of the box section, each 160mm long and welded them together.

Flatness is important here! Before welding them, I put them on a massive piece of half-inch-thick steel plate, clamped them down with three clamps, then clamped them together with two more for good measure. This eliminates any possibility of heat distortion on several axes. It was probably overkill, but those things were kicking around the workshop and I might as well use them.

I had to grind these welds flat later; this will be going into the mill for drilling.

Next, I marked up and punched holes. Two of these are where the shock absorber mount will go through.

Here's the basic plan:

And here's something I threw together in FreeCAD to maybe better show you what I mean:

The big hole in the center is to form the curve where the perch welds to the axle tubing. It will be drilled with its centre right at the join of the two pieces of metal. That is offset from the left edge by 63.5 millimeters. I measured the correct position for the two 19 millimeter holes (B and C in the photographs) as 31.8mm from the outside top and bottom edges, and 140mm from the left edge. (And if you are surprised at the decimal precision: I used a digital vernier for most measurements, and I don't believe in throwing away precision until the very last moment.)

The idea for the big hole was to send a 76mm (3 inch, which 75mm is not) TCT hole saw in a single pass. Of course, if you look at this photo of what hole-sawing removed from this...

...you'll quickly work out why that could not possibly work (hint: hole saws are not arbitrarily deep). Instead, this had to be done in more than one pass. First, the TCT started the cut, and went down as deep as that could. Then a somewhat-deeper Starrett hole saw (FCH0300) followed it to extend the cut deeper. I actually bought the Starrett one as a backup to the no-name TCT in case the latter exploded; I wouldn't have been able to complete this mini-project without it.

Then, without moving the workpiece on the bed of the mill, the hole saw was swapped for a long 6mm drill bit. That went through all the way to give a marker on the other side, so that the workpiece could be flipped over and the hole saws sent through again, using the hole from the 6mm bit as a guide. That still didn't cut all the way through, but the remainder was so small that the above piece could be knocked out with a hammer.

The 19mm holes were made in the places they should be made in the obvious way, and I shoved some correctly-sized tubing through them them. This was 19.05mm tubing (it's fine, it went through with only a little persuasion), with 3.25mm walls, which makes it just the perfect size on the inside (12.55mm) to shove an M12 bolt through it.

I needed to reproduce something like the donuts that go over the bottom of the mounts. I could have turned these up on the lathe, but it was less effort to just cut off the originals. That means that the rear axle on the P5 will have at least one actual Rover part!

The donut I wanted is not welded to the sticky-out bit, but it might as well have been. It took some bribes with the oxy propane torch (yay dangerous toys) and a big hammer to get them separated.

With that all welded up at the end of day 1, it looked like this.

That's fundamentally complete! You'll notice that I haven't separated them yet. This is not an accident. I take heat distortion much more seriously than I was this time last year. Welding generates lots of heat, and angle grinders generate somewhat less. Keeping them welded together should mean they reinforce each other against it. I didn't separate them until there was nothing more I could do without separating them, and after they were separated I didn't cut the bottoms of the boxes out until there was nothing left to do.

There was one more hole to drill: the locating hole for the leaf spring. That is on the top, in the center on the left-right axis, and at the same 63.5mm front to back (offset from the opposite end to the shock absorber mount). This needed a 29mm hole. I made that hole look nice with a big (40mm) countersink .That might not be necessary, but the originals had it and it looks nice.

I could have just separated them here and called it done. But I wanted these to look right, so I spent a few more hours with a grinder on day 2 to shape them and tweak them till they looked nice (and somewhat like the originals, with the help of a marker pen and a crude cardboard template). And by 3pm, even factoring in waking up an hour late because of the clocks changing...

...I had some Rover P5 leaf spring perches, ready to be welded on to a completely different axle housing. I am more than happy with how those turned out. I wouldn't look like a real fabricator if you stood back 20 feet and squinted, but this is definitely a nice accomplishment for someone who has no actual clue what he is doing.

See you next time. :)

Thanks to Maurice for the help working the mill and for some quick thinking that saved me at least an entire day!

I don't like working with 304 stainless

Many tools died to bring you this information.

Overkilling the Rover P5, part 2: the gearbox

I know what you're thinking. It's something like "Lewis, a Rover P5's prehistoric manual gearbox will explode if you put 350 horsepower through it". And you're right!

Here was my gearbox. It's attached to my old engine here; I didn't separate them for the eBay auction because I couldn't be bothered. It was an ancient 4-speed manual. It is so old that it actually pre-dates the Rover P5; it's similar enough to the seventy-plus-year-old Rover P4's gearbox that most of the innards are allegedly interchangeable.

Over to the right of the engine and gearbox is a Laycock overdrive, which gives it a fifth gear for cruising.

The overdrive is high-tech for the 1960s; it's electrically-operated via a switch on the dash and had a lockout which stops it being used in fourth gear.

It's an amazing thing! It is far removed from the primitive Fairey units of the era, which used a mechanical lever to engage it, and an instructional plate in the interior telling you what gears you should use it in.


And before we go on, let's have a little digression back in time to "Lewis removes the old engine and gearbox".

To get the gearbox out, you must first remove the overdrive. To remove the overdrive, there is a ring of bolts, four or so of which can only be accessed through a small hole in the transmission tunnel. It's fiddly, it's time consuming, but it's doable - just one of those silly fiddly jobs you get on any car of any age.

Alternatively you could, as someone in this car's past decided, just fucking slice open the transmission tunnel, peel it back, and fold it back into place when you are done and not even bother welding it up because you're time constrained and nobody's going to notice and you are an idiot mechanic.

All I'm getting at, is that if anyone worried that I'm ruining a perfectly good P5 Mark 1A export CKD: I'm not. :)


Anyway, I no longer own any of that. The engine and box sold together on eBay. It made little money. I am comfortable with that fact, because it is going on to a new life in a Rover P4. It will be a substantial upgrade in the much-lighter P4 - that'll be a 115 horsepower engine in a 1400 kg car. Nice!

The transmission wasn't any good to me, because even if it could take the power I want to put through it (it absolutely will not), and someone actually made an adapter plate for it (nobody does), I wouldn't want to use it for a bunch of reasons, one of which is that parts are very scarce and expensive these days. Instead, I ended up with...

...a BMW GS6-53DZ HGD six-speed manual gearbox from a BMW 525D, and some shiny things. How did I get here?

Above all, I wanted a manual transmission. I don't intend to ever buy a car with an automatic transmission, and I'm definitely not going to build one. If I liked automatic transmissions I'd have lots of bolt-on options, some of them quite cheap, but I don't, so I don't.

With auto boxes written off, the bolt-on options (other than some mad Quaife sequential and yes I did seriously consider this) are exactly two: the five-speed Tremec TKX and the six-speed Tremec T56. You can probably find a T56 at a scrapyard, if you live in the United States where the LS engine was born and cars with LS engines were sold by the millions. That is not where I am from, and a T56 is a specialist item over here. You can get them new in the UK, but they run about £4000 before you've bought a bellhousing for it. The TKX is only slightly cheaper.

I'm not building this car on the smallest possible budget. That ship sailed two seconds after I decided to do an LS swap in the UK. But if I can save several grand without compromising anything I will, because that gives me more money to spend on takeaways and more car parts.

So, the BMW gearbox. While I don't want a four-speed gearbox, I have no particular desire for a six-speed gearbox; I am not building this car for motorway-mile high-scores done quietly and economically. I bought this because it is a reputedly strong gearbox. It is rated for 391 lb-ft of torque (or 530 newton metres, the "53" in the name). Most likely the GS6-53DZ is good for much more torque than that (evidenced by the drifters routinely not blowing them up). I should have headroom if I do something silly like bolting a turbo to the LS1 which absolutely is not occupying my every waking thought right now.

As well as being a strong gearbox, it is a strong gearbox I can get cheaply. The HGD-suffix gearboxes aren't as plentiful as the others, but they're not hard to find for less than £500. That means I can justify buying a spare, for when I inevitably destroy it doing massive smoky burnouts. And if anything does go wrong, it's not hard to get parts for them nor to find anyone to work on them in the UK.

To get any of this to work with the LS engine, I got the PMC Motorsport adapter kit. Those are the three shiny bits you see in the photo above. To the left is a gorgeously-machined plate which bolts to the BMW gearbox. On the right is another gorgeously-machined plate which bolts to the LS engine. These two bolt together to mate the engine to the gearbox. In the middle is a custom flywheel, which bolts to the LS1's flex plate/starter ring.

I will need to make a minor modification to the gearbox (the guide tube needs shortening), and there are more parts to be acquired before this is a working transmission setup, but I'll deal with that later. Let's get out the trusty caravan stands and transmission jack and find out how this gearbox doesn't fit the car, and what we can do about that fact.

The good news is that the body of the gearbox fits within the subframe! My fear was that I might have to cut out and reshape the rear crossmember of the subframe to do this, but that isn't necessary. The body of the gearbox clears every part of the subframe, and the output spider clears the rear of the subframe. It's tight, but that's another way to spell "fits"; nothing I actually need hits anything that can't be modified.

The BMW OEM rear mounts are entirely negotiable. It's a three-part system which is very light, but also very large.

There's no room to fit it behind the gearbox in the Rover P5 subframe. This would require cutting out part of the rear-most crossmember - a thing I wish to not do. And even if I could use it, it has rather too much flex for my tastes. It makes sense for minimising noise and vibration in a daily-driven car, and does not make sense for me.

In designing a new mounting, it would be really easy to design something with the access I have now, which would be impossible to fit it when it was on the car. For example, the four bolts on the back of the gearbox are really easy to get to - when I'm working with a subframe sat on caravan stands in my workshop. They're a lot more fiddly to get to from underneath, which is how I will be accessing them when fitting the gearbox to a subframe bolted to the car.

So with some thought (and a stroke of genius from Maurice), I came up with a two-part mounting which is a lot simpler and stiffer than the OEM mounting. Here is part 1.

I copied the hole spacing and the vague shape from the original BMW gearbox mounting and traced it onto 3mm steel plate. I got the holes not quite right the first time; I needed to relocate one of them further out-board, and rather than move that one out all the way I decided to relocate all of them only a little to compensate, and reinforce those areas with welded-on washers.

On that plate is a piece of 28mm ID steel tube, separated from the plate with some random bit of angle iron I had kicking around (probably 20mm). That isn't really there for spacing. Instead, it made it much easier to get the tube aligned perfectly level (± my sloppiness) when I stuck all of this together.

This bolts to the back of the gearbox, and will be bolted to the gearbox when I fit it into the car - solving the problem of those four bolts being fiddly to get to from underneath.

The other half of this re-uses the original two holes in the subframe for the rearmost gearbox mounting bracket on the P5's subframe (but I drilled these out to accept an M8 bolt, only a fraction larger than the original Imperial 5/16). And I took the original P5 rear mounting bracket...

...and Frankensteined it with more 3mm plate into whatever the fuck this is:

These two parts join together via a generic 90mm Powerflex top hat bush which I actually bought for a different sub-project on the car (tell me if you guessed what that might be). When it all bolts together, it looks like this:

This, I think, is a rather neat solution to the problem. It fits. It's simple. It'll actually be accessible in situ. And it'll give me enough flex to stop the vibration driving me nuts (and hopefully enough to stop the gearbox from shaking itself apart), unlike a solid mount.

And with everything else bolted up...

...I get another another milestone: a (stunt) Chevy engine and BMW gearbox bolted into a Rover subframe with no outside support.

And that, is enough car for now. :) See you next time.


Part numbers from this post:

  • Gearbox: BMW GS6-53DZ HGD / 23 00 7 522 205 (ZF 1067 401 060)
  • Gearbox adapter: PMC A-LS-HGD-240
  • Bushing: Powerflex PF99-101
  • Stunt engine: Speedway Motors 12640748
  • Transmission jack: Draper 09021
  • Steel from Thomas B. Bonnett

The one where Lewis totally, irreversibly loses the plot

This was my Rover P5's engine.

You'll notice that it's not in the Rover P5.

So, the workshop build was completed, giving me a much nicer space to work in. The Mazda 323 GTX was fully someone else's problem while the engine is (still) being rebuilt. And some time ago I made some major life changes to give me more time and motivation; it worked! Which meant I could finally get back to the Rover P5, for which there are thousands of pounds worth of parts in storage to fit, and a list of jobs that'll keep me out of trouble for the foreseeable future. And I found myself procrastinating, again. Why?

Someone wise once told me that when I find myself procrastinating, I need to look for the fear. I think one of those fears is that I have seen some really nicely-done Rover P5s, and I doubt I will ever be able to do one as well as those. There's a fear of being judged by all the people that build all those perfect P5s. Half of that doesn't matter (it doesn't need to be perfect) and the other half is mostly something I made up (there are so very few judgy people in the classic car world and I'm actually thick-skinned faced with actual criticism rather than imaginary criticism anyway).

What I also found out was: When finished, I just wouldn't find the P5 fun to drive.

It's a beautiful car, maybe one of the most beautiful cars ever made. This one in particular, in this colour, is so beautiful that I was forced to impulse-purchase it. And yet, it has an engine designed shortly after World War II, a carburettor, a distributor with points, and 115 horsepower. The horsepower number sounds OK, but it is in a car weighing 1700 kilograms. It is substantially slower than my totally-standard 1.4 litre Ford Fiesta from 2004, even if you forget the fact that my Fiesta can corner and a P5 can't. Or that my Fiesta has actually worked every day, where a car with a carb and a distributor with points is a lottery.

A 3-litre Rover P5 is not fun to me, and I live my life one fun at a time, or something.

Anyone sensible would have sold the car at that admission-to-self. Actually, anybody sensible would not have bought impulse-purchased a non-working car for £3500 unseen over the Internet, so wouldn't reach this point in the first place. Because I'm not sensible, I decided to start a new plan: make a Rover P5 fun to drive.

I actually had several ideas. One was to develop fuel injection & electronically-controlled ignition for the IOE engine, to at least make it reliable and probably get a few more horsepowers out of it. That would open up more options, like shoving a massive turbo onto it for the lolz, and well, if that idea sounds insane that's why I didn't do it.

So, what else? Rover V8? 2JZ? K20???

They all have their merits. And yet, those are all petrol engines. Any discussion of drivetrain swaps in classic cars must eventually face the spectre of electrification.

Electric means 100% torque at zero RPM, fewer moving parts, near-silent; it's, like, the future, man. And though some consider it sacrilege to electrify a classic car, it can be argued that changing an old, unreliable, underpowered petrol engine to a modern, reliable, much more powerful one is changing the character of a car every bit as much as electrifying a car is. And nobody but the dullest of rivet-counters objects much to all the mad engine swaps out there; why object, then, to electrification?

But there's more. We all know that petrol engines pollute. Some would argue that, whatever the historic value of these old cars, an internal combustion engine in a classic car in The Current Year is like painting an historic house with lead paint, heating it with trees felled from old-growth forests, and emptying a chamber pot from its windows onto the street. Such a thing may have been a grim necessity in the past; doing it today when we have clearly better and cleaner options is entirely irresponsible. And, honestly, it's very hard to argue with that.

And that is why I bought

a Corvette engine, because fuck 'em.

This is the Chevrolet LS1, a 5.7 litre V8 petrol engine with about 350 horsepower and over 350 lbs/ft of torque, and an aftermarket that will take them to arbitrarily large power numbers. And it makes cool V8 noises! Now we're talking!!

This is not another dumb impulse purchase (he said with a straight face) but the product of a lot of thinking. I wanted reliable power in a compact package that had some hope of fitting in the P5.

I needed something well-documented; "LS1 into a Rover P5" might not have been done before, but "LS engine into something that shouldn't have an LS engine" has been done probably thousands of times. And anything you are thinking of doing with an LS1, the fine folks over at the LS1Tech.com forums have probably already done it five times and found out which of those five ways works best and which four don't.

I needed something somewhat modern; no LS engine has a distributor, and they definitely do not have a carb, which would allow me to preserve my unbroken 41-year streak of not dealing with carbs. (Yes, I would rather re-engineer a car entirely rather than ever deal with bloody carbs and this is literally what is happening here.)

It is a simple engine, compared to other modern engines with the same power figures. I reckon I could hold one in my head. Simplicity means less to go wrong, and simplicity means that if it does go wrong it's straightforward to fix. Simplicity also means that they could build them cheaper, and make up the difference with displacement.

It's kind of cheap. There are no scrapyard LS engines in England; this cost me nearly £4000 and it had to be shipped from Ireland. But if you price it against any other 350 horsepower engine it's not that bad. Sure, everyone's getting 350+ out of their K20s these days - but you won't get a K20 pushing out that kind of power reliably for £4k. Or maybe I could probably pick up one of the German V8s for that money - but then that would be a monstrously complicated engine with no substantial after-market. Or I could pick up a JZ - for substantially more money, and complexity, and that still wouldn't push 350 horsepower without modification. (This was actually my original plan. My brother talked me out of it. I am glad he did, and not just because that means losing the next several years of my life to this is not at all my fault.)

I used to hate the saying there's no replacement for displacement. You probably won't find me using it again after this very paragraph, in fact. But I have to admit there's something to it. For reliable power at a given price point, sometimes it's best to throw a bunch of unsophisticated cubic centimeters at the problem. "Quantity has a quality all its own", as Stalin once said while trying to find positive things to say about my welding. Or, "when in doubt, use brute force".

And that is how Lewis convinced himself that buying a 5.7 litre V8 was a good idea! So...

Does it fit?!

As if I'd ask that sort of question before spending all of my money on an engine! But I had a hunch that it would. Its 5.7 litre displacement seems ludicrous to anyone whose patriotism includes no eagles, but it's actually a remarkably compact engine. It's substantially smaller in most dimensions than a Rover V8, and we know a Rover V8 fits in a Rover P5. But measurements can lie; I wouldn't know for a fact that it fits until I shoved it in the engine bay and saw what it collided with.

What I didn't feel like doing, though, was lifting 215 kilos in and out more than once, because setting up the engine crane every time would be effort. Which is why I bought another LS engine!

The eagle-eyed among you will notice that this isn't really an LS engine. It's the wrong colour! This is, instead, a mockup engine, or as I like to call it, a stunt engine. It replicates all of the critical dimensions and mounting points of a real LS engine. My one even has fake cylinder heads! And it weighs like 8kg, or 15kg with the "heads", which is a weight nobody minds lifting in and out of a car. You can get one at Speedway Motors in the US. (Their website plain doesn't work with non-US delivery addresses and telephone numbers though; you'll have to give them a phone call if you're outside the US, and brace yourself for the shipping cost.)

So, I dropped the stunt engine into the P5's engine bay, in about the same place that the original IOE engine sat.

Let's talk first about what "fits" means. What that really means is that it doesn't hit anything that I want it to clear. There are some things that are non-negotiable. It can't interfere with any part of the steering. I don't want to chop out the front crossmember. I don't want to cut into the bulkhead. I don't want to redesign the suspension. I don't want to have a bulge in the bonnet, because that would look silly. Anything else - especially adding metal to the car - is up for grabs.

As you can see, there's miles of vertical space there, and no clearance problems around the heads. Let's look at the air conditioning pump.

Acres of room there, because I threw the air-con pump onto the scrap pile (not gonna need it, and saves 8kg), so there's nothing to clear. Woohoo!!

Let's look at the other side.

The big mount on the right-hand side is one of my old engine mounting points; that's going to get chopped off. The small brackets at the front are for the radiator or something, I forgot. Don't worry about those. They're in the way, but they can be anywhere on that front crossmember, and anyway nobody said that the radiator must be at the front of the car.

:)

Still, there's a possibility that the alternator will not clear the front crossmember. It's going to be tight. Even if it fits, it might be a pain to get in and out. It might necessitate a more compact alternator, and if that doesn't work I can mount the alternator at the top right of the engine as the LS-engined trucks have it. There's some possible solution for it out there even if I don't know what that is yet, so let's call that OK.

It's a bit tight around the steering box and somewhat less tight on the steering relay on the other side. Tight, but some design of engine mounting won't conflict. I can fabricate around this - run mounting rails from the rear to the front - in a way that'll allow me to get the steering components in and out.

And there's enough room at the back that I'll be able to unbolt the engine from the box with less hassle than the original engine! But while we're looking at the back...

No possible design of sump is going to clear that rear crossmember. This crossmember is negotiable though; it needs to be more or less there, but it doesn't have to look like this. It can plausibly be dropped lower and I'll pretend that this isn't going to result in any ground clearance issues.

My brother has a way of asking the obvious questions that I did not think of, such as...

And well, yes, I can.


So, an LS engine fits in a Rover P5, for some value of "LS engine" and "fits". With that point proved, I could drop the front subframe.

This was easier than expected. The six bolts that held it in actually came undone, perhaps because of plenty of penetrant applied in the days before. The only other things in the way were the old over-complicated handbrake mechanism (which is going to be replaced by something much simpler when that comes around), an earth strap, and a brake line, all of which were swiftly defeated by an angle grinder.

With the subframe on the floor, I could take a look at the original engine mounts. I said "Corvette engine" earlier because it is, but it was also fitted in many other cars, and I think these are Pontiac GTO mounts (which obviously implies this engine came from a GTO, and not a Corvette). With those bolted to the stunt engine, I could see that re-using them is somewhat plausible.

But that is without the engine at the correct height relative to the subframe; they're both sitting on the ground.

What is the correct height and location of the engine, though? I actually decided on this early on, before the old engine came out. It was to copy the forwardmost point and the crank centreline of the P5's engine. The engine is exactly central in the subframe. The very front of the vertical centre of the front pulley is 160mm from the deepest point of the dip of the front crossmember, with the front face of it sitting exactly in the fore-aft centre of that crossmember. The engine tilts back at an angle of 3 degrees. That, I decided, will be the position of the new engine; it is as good as any.

But to get the stunt engine in position on the subframe, how am I going to measure those dimensions relative to the front face and centreline of the front pulley, when my stunt engine does not have a front pulley? The answer, obviously, is to fit a front pulley to it. Here's what that looks like.

The eagle-eyed among you will notice that this isn't really an LS1 front pulley. It's the wrong colour! This is, instead, a mockup pulley, or as I like to call it, a stunt pulley. I don't actually need a whole front pulley; I need something to mark out the centreline of the crank and the stick-out of the real pulley. So I welded some shit together and came up with this. The centre of the rectangular cut-out on the front of the stunt engine corresponds with the crank centreline of the LS. After I positioned it as precisely as I reasonably could, I marked some sharp lines with a paint pen on the stunt pulley. When all four of these markings are lined up with the edges of the rectangular cut-out, like this...

...I know that the stunt pulley is exactly where it needs to be. This saves a lot of tedious measuring (it took 15 minutes to line this up perfectly to make the markings) and it's easy to validate with an eyeball that it hasn't shifted anywhere.

Anyway, let's get the fake pulley onto the fake engine, put it on the subframe and see how it goes.

To do this, I bought these cheap adjustable caravan stands to hold the subframe up to a working height...

...and lifted the stunt engine up to the correct height with this jack designed for gearboxes (which means it has a perfectly linear upwards lift unlike a conventional jack)...

...and with everything in more-or-less the right place....

...bolt the steering box and relay into the subframe and see what fouls what.

It turned out I can't use those GTO engine mounts, because they foul on the steering arms - even if they could bolt to mid-air there would be no clearance there.

Shame, because I liked them, and didn't really want to make my own, but needs must. I made an initial version of my mounts out of cardboard.

Which was fine, except there was no space to get a bolt into the engine block, so I came up with a second one...

...which I could fit to the stunt engine block using a bolt. Which meant I could turn that into steel, and to talk about turning it into steel we shall have a brief diversion about tools.


Tools! I like them! Tools help you make things! And yet I am prone as anyone to thinking that just one more tool is going to solve all my problems. What I actually needed to do instead of buying more tools was to use the ones I have.

Some of those tools are three 115mm angle grinders. They all perform exactly the same. I like having one with a cutting disc, one with a flappy wheel disc, and one with a grinding disc. It saves having to change discs, which I don't like doing because life is short.

The other one is a cheap MIG welder. It's a Clarke 151EN which my brother gave me for free (thanks Alex), which he bought for about £200 in 2013 and didn't use much.

I've converted it to use a Euro torch, to make it easier to find consumables. I also ripped out the gas feed from the welder to use it solely with flux-core (gasless) wire. A lot of people don't like flux-core welding. I kinda like it; I get nearly the same results and I don't need to worry about running out of gas. It's easy to get flux core wire from Amazon; it's more effort to get a gas bottle refilled. And I want to prove a point, and only to myself: that tooling is not what is coming between me and getting actual stuff done.

It's a cheap welder so I might burn it out before I've finished the subframe. Or I'll end up converting it to gas! But for now this will do.


Anyway, with the cheap welder and a selection of angle grinders I turned my cardboard engine mount into 3mm steel...

...and then remade it again with a 6mm main plate chopped off a generic LS engine mount, and 3mm steel for the rest...

...which almost solved the problems I had from the first round. (I'll still need to fettle it in the press a bit to get rid of some heat distortion from welding, and tidy up the welds; that can be done later.)

Those with some the mounting rubbers clear the steering arms with room to spare.

Nice! And I'll need that room to spare, because I need to fabricate some rails for these mounts to sit on. And that starts with these:

These took three revisions across three days to make. Someone who is good at this would have done it in a fraction of the time; I don't know what I'm doing, so it takes longer. That's part of the fun of this. I've never done an engine swap (if you don't count putting a slightly larger-displacement version of the same engine into a car), let alone an LS swap into something that never had an LS. I can chop up metal with an angle grinder and do an ugly job of sticking it together with a gasless MIG I got for free; those are my "skills". And this rules! If I only did what I know I'd still be soiling myself and crying for food. Just give it a go; what's the worst that can happen?

Anyway, the idea behind these sections of rail is that I can bolt them to the engine mount. Then I can get the subframe exactly level, and then get the stunt engine in exactly the right position relative to the subframe (rather than the crude approximation I did to check my steering clearances). This took three hours! And it required continued adjustment over the course of the day as the temperature changed and when I accidentally kicked the caravan stands. From there, you can fabricate outwards from the stub rails towards the front corners of the subframe. This is how a partly-completed rail looks:

And getting the engine in the right position gets a lot easier after the first quarter of these rails is in place! Here is how it extends backwards:

The box to which it extends rearwards (right in the pic above) is not strong enough for these purposes. It is a fairly-thin reinforcer between the suspension arm mounting points and the rear crossmember, which are strong enough for this job. I'll reinforce this box later.

Anyway, spend nine days straight on doing all of those things and you get something that looks like this:

That's close enough to done for the moment; it holds the engine in the correct place and at the correct angles for us to move on for a while. Let's take a look at that alternator clearance again.

That nearly fits. Nearly fitting means "doesn't fit". But "nearly" also implies a couple of solutions that aren't top-mounting the alternator (which I don't want to do because this will merely move the clearance problem somewhere else rather than solve it). One of those would be to simply notch the new subframe rails a little. But I think it would be better to solve this by fitting a modern, more compact alternator. That's the obvious solution to me, because I want to replace my alternator anyway; it and all the other ancillaries on the engine came without no guarantee that they work.

Let's talk about the sump for a second. It's abundantly clear to me that the original front-sump would never clear the steering gear. So I bought this Camaro rear-ish sump instead...

...which still won't clear the steering gear at the front! But it's a better starting point because it has the oil filter at the rear. That means it can be modified to fit. What I knew all along was that the rear of any LS sump won't clear the centre crossmember. There was a little clearance on a sump-less stunt engine - in fact, millimetres from clearing the centre crossmember with the rear sump. But modifying the sump at the rear is not a solution. The sump on the LS engines are structural; the bellhousing bolts to it. So there's a hard limit on how shallow you can make a sump at the rear, and I think the Camaro sump is as shallow as it can be in the areas that interfere with the crossmember.

Instead, that means that we have to brace the chassis with some angle iron, cut the centre crossmember out...

...and remake it, but about an inch or two lower. To do this, I took an oversized length piece of 25x50x3mm box section steel, and clamped that to some temporary angle iron pointing downwards from the outside of the chassis rail, spending as long as I needed to get it in exactly the right position.

From there, I could add some steel plate to hold it in place...

...and then add a whole load more steel and miles of flux core welding wire to make it into a centre crossmember, and by the way there goes another three days.

At which point I started to wonder how much of the weight savings from fitting an LS1 are being negated by the sheer volume of steel I'm putting into this!

After that, I knew that there was going to be some clearance problems around the starter motor. So I cut out some of the steel I just added, chopped a slice of 100mm steel pipe, and shoved that in the hole I made (it's just right of the rear of the engine below).

As promised earlier, I reinforced the box between the suspension arm mounts and the stub of the old rear crossmember by adding even more 3mm steel.

And with all that done, I could spend a day and a half going around fixing up my bad welds, cleaning up a little,...and add a little zinc-rich primer as a temporary to stop everything from rusting faster than a British car leaving the factory. While Chip Foose is not going to start hammering my inbox trying to hire me for my beautiful fabrication skills, it looks pretty OK for someone who has absolutely no clue what he is doing.


So that's where I am now: an LS engine (or stunt version thereof), sitting in a Rover P5 subframe, with most of the things clearing the things they need to clear. Some things need fettling, and some things might still need to be welded to it; this is not its final form. But this is done enough that I can call it done for now.

And that, is the easy part out of the way. :) See you next time.


Part numbers from this post:

  • Stunt engine: Speedway Motors SoloSwap 12640748
  • Transmission jack: Draper 09021
  • Welder: Clarke 151EN
  • Sump: General Motors 12640748
  • Die grinder: Sealey SA671
  • Zinc primer: U-POL WELD/AL
  • Cardboard: Amazon single-skin envelopes (from my deliveries), Felix cat food boxes
  • Steel supplied by Thomas B. Bonnett

A fully-dressed Chevy LS1 engine weighs 215 kilograms, or 474lbs

A fully-dressed LS1 V8 engine weighs exactly 215 kilos. That is 474 Freedoms if you are from the United States of America. I put it in the title of this post to save the world a click.

This includes everything, other than a transmission, which is normally directly attached to the engine - power steering & air con pump, alternator, inlet & exhaust manifolds, flex plate, engine mounts, fuel rail, engine side of the wiring loom, etc. There's plenty of numbers out there on the Internet that involve the words "about" and "I think", and I didn't like that, so I lifted one with a £35 crane scale from Amazon to find out for sure.

You're welcome!

I know what you're thinking. It's "how would you know what I'm thinking?". But for the sake of the narrative I'll say that what you're actually thinking is "Lewis, why on earth would you be weighing a Corvette engine?"

That's a good question! And well, things just escalated, a little...

Let's play: "remove suspension bushes", on easy mode

This is a top link from the front suspension of a Rover P5.

This is part of the "rebuild the entire front end" subproject of my Rover P5 project. Among other things, I am replacing every suspension bush. That requires removing perished six-thousand-year-old rubber suspension bushes. This is nobody's favourite job, for any person that is lacking (say) a 12-ton hydraulic press.

I learned a little trick that makes it painless to remove suspension bushes. It uses tools and random junk which you probably already own if you have a car project.

First, identify the side of your suspension component that has more protruding bush than the other. We're going to call this side the receiving end.

If neither is an obvious candidate, then pick either one. It's not hugely important; it means marginally less effort to remove it.

Next, find a socket that fits fairly snugly over the bush on the receiving end. You should find one just large enough to accommodate the bush inside. The depth of the socket is not too critical.

In this case, a 32mm Metrinch socket was a perfect fit.

Next, find a bolt. It should be of the right diameter to fit through your socket. It should have sufficient length to go through the socket, your suspension component, and leave enough thread poking out the other end to accommodate a nut and a washer. 10 mm of "spare" thread (that's 3/8" if you're reading this in the United States or in 1965). Slide this and a protective washer through your socket.

I did not have a bolt of just the right size, but I did have a random length of somewhat-munted threaded rod which was of just the right size, so I used that instead.

Next, put a nut and washer on the other side of your bolt.

Ideally, the washer should be just a little smaller than the internal diameter of your component. It is sufficient to be just somewhat larger than the nut.

Next, put a spanner on the bolt head on the receiving end, then tighten the nut on the other end. You can do this by hand if you have to. I had other things I wanted to do that day, and I also have an impact wrench, so I went full send with an impact wrench.

This will extract the suspension bush into your receiving socket! ⭐ 🎵 IT'S A KIND OF MAGIC 🎵 ⭐

And the now-thoroughly-destroyed bush should quite easily pull out of your socket.

Do that another three times, then spend a couple of afternoons cleaning them up, and you'll end up with four spiffy-looking suspension top links...

...which do not have suspension bushes yet! Most people working on most project cars would be happy to fit polyurethane bushes from here. Polybushes are usually easy to fit without a press, so the next step would be straightforward and would not require a press. I would not want polybushes on the P5 even if such a thing existed for the P5; I want rubber bushes to keep the original characteristics of the car. I don't trust this same method to fit rubber suspension bushes (after all, I didn't care about destroying them when removing them), so this meant ordering that 12-ton press I mentioned. Oh well, I'm not convinced a press would have done the job any faster.

There are other field-expedient ways of removing bushes. You can burn them out with a blowtorch, but this stinks and takes forever. You can use a hole saw of exactly the right size. That stinks slightly less but also takes forever. This method is not just smell-free, but it is also very quick; it took about 20 minutes to find exactly the right combination of random bits of stuff in a box of shit and about 2 minutes in total to extract all four bushes. I hope this helps someone else.

Experiments within a project within a project

This is my brake servo.

I removed it from the car as part of the "refurbish the brakes" subproject of the "financially ruinous Rover P5" project. It's a separate unit from the pedal-activated cylinder unlike it is on almost any other car, because Rover loved overcomplicating things back then.

The vacuum chamber (the large cylinder on the right) was half-full with brake fluid. As the name "vacuum" implies, there is not meant to be any fluid in there at all. That was quite the surprise when I took it apart, and also a surprise which caused me to lose a full set of clothing. There was little way the servo could have been functioning, and I suspect that aeration of the brake fluid would have caused the braking system to be minimally-functional even without the servo's assistance.

I suspected this seal may have been the problem:

But, the unit was so ugly that I cannot say for sure that any single component was at fault. The photos you see here were after a blast of electrical contact cleaner (which is great for cleaning up anything that isn't an electrical contact).

I take a lot of photos in the course of doing things as reference for how things need to go back together. In the above one, you can see quite a lot of mank in the servo's internals.

Under other circumstances, I would have considered giving it a wire brush on the outside, a soak in petrol and a skim with very mild abrasives to clean the cylinders. But while I was ordering an electroplating kit for the various fasteners I have removed from the car, it seemed a good idea to order an electrolytic cleaning kit` as an experiment! As if I didn't have enough new things to learn...

Electrolytic cleaning is, in my head, electroplating in reverse. Rather than using an electric current to attract material onto a thing you want to plate, you operate it in reverse so an electric current attracts material away from a thing you want to clean.

It works quite well. In the process of working quite well, it turns the electrolyte (this is water mixed with powdered sodium carbonate) into the only thing I would like to drink less than brake fluid or Foster's.

The exterior of the servo still required some love from a wire brush on the outside after the cleaning. The two bores feel very nice after the cleaning, so I will leave them be.

And the part that makes me happiest, is that the Girling wordmark is now visible!

And that...means that after two weekends I am half-way through rebuilding one of the components of one of the sub-projects of the P5.

Onwards!

The little things

I love good documentation.

The Rover P5's workshop manual - the original one written by the engineers at Rover - is the best manual I have ever used. I especially enjoy that the manual sets out exactly what tools you will need including the sizes of spanners in the preface. This is the only workshop manual I have seen that does this, and it makes every job a tiny bit faster and a tiny bit more fun.

How many times did I work on the Mazda and read through a 20-step walkthrough and then in step 19 it says "use special tool, part number F-U, which is obviously completely unavailable for any amount of money to remove this delicate thing that will definitely break if you try to use cruder methods to remove it"?

Good documentation is important, and I believe in thanking people for good work even if I am late to their particular party. So, thank you, Rover engineers from 60 years ago. You did a fantastic job.

In which I explore just how bad a decision this P5 was

On the weekend that has most recently passed, my mission was to find out exactly how bad or good Penny the P5 is. The good news is that everything is very good and really solid, despite the scruffy appearance. South Africa was very kind to Penny. There's only one little bit that looks a bit sketchy on one of the front inner wings and where it joins to the sill.

Also, there's almost nothing missing. Those bits that are missing might be a pain to find, but none of them would stop me getting it running. (I was immediately 250 quid poorer from buying three of those bits on Saturday. It's only money...)

What is not missing is this. Behold, 3 litres, 6 cylinders and 115 horsepower of IOE Rover FURY:

If you look at the far bottom right you can see a glass washer bottle! I was last-Saturday-years-old when I learned that was a thing.

I do wonder how necessary that gigantic airbox is. I also wonder if it could be replaced with a tiny K&N or Mishimoto cone filter. Or maybe a Range Rover airbox, since that's what I know to be sensibly-sized and what I know to have very little intake noise. Or maybe it can wait till I pull off my next terrible idea!

So, back to what I should be doing, which is to get the engine started. I said there's nothing that would stop me getting it running. But it's clear that all the wiring is not in a great state.

The wires exist and have not set on fire yet. It will almost certainly work to whatever extent it can, until it sets on fire. I can certainly foresee a situation in which I am going to replace some substantial part of the wiring just to get the engine started...and if I'm going to do that, then I might as well go all the way...

...which is to say I have decided to rewire the entire car, and do so for negative earth & an alternator, before I even attempt to get the engine running. This might be a bit of a brave move given that my knowledge of electrical circuits is limited to basic knowledge of how DC circuits work. Oh well, as a wise man said one time, the best way of getting started with something you don't know how to do is by starting to do it. Fortunately, you can buy entire wiring harnesses off the shelf from Autosparks, which should be about £512.40 if you add in the optional wiring for an alternator, electric fan, and radio feed (I don't intend to fit a radio to it, but I don't want to rule it out). Yep, it's only money...

As I knew before I bought it, the interior is totally shot:

It's not just the cover that is torn; whatever previously resembled foam on the seats has turned into something that is not at all like foam. It has the texture of compressed sand, in that it crumbles into dust when you squeeze it in your fingers. Also, almost every piece of woodwork has delaminated, and those bits of wood that are covered in other materials have had their coverings disintegrate. I think I might be able to learn enough woodwork by myself to fix some of those bits. The upholstery almost certainly needs to be done by someone else.

So, this isn't going to be an easy recommisioning - not that I thought it would be, even before I introduced this new complication. And definitely not cheap, though I knew that would not be the case either. I can afford it, but I will say that I will definitely be eating a lot of instant noodles over the next few months...

...and these are the God-tier mother of all instant noodles. Something, almost certainly, for a future review on the very website you are reading now.

Onwards!

Introducing Project Penny

I've been thinking about a backup car for Amy, my Mazda 323 GTX, for quite a while now. Not because I don't like her. She rules. She also doesn't work all the time. I've been wanting to take her off the road and get her indoors for a while now. I'd like to modernise & rationalise the turbo setup and the engine management. I'd also like to get her bodywork sorted once and for all, rather than the remedial work that I am doing now.

And also...a tiny, ridiculously rare turbocharged hatchback with poverty-spec features (wind-down windows, no air conditioning, no safety features at all) isn't exactly what I need in a daily. I'd actually like to make it to destinations, and get there in comfort, without spending three million pounds on fuel.

So. I needed something cheap, economical, sensible, modern, reliable, easy to find parts for, and with at least a few modern creature comforts as well. And that is why I bought

a three-litre Rover P5 from 1964, which might work, because Lewis Logic fucking rules. Say hello to Penny the P5!

(The car purchase is real, but the narrative may or may not be. All I'll say, is that when I first started writing this the narrative hook was along the lines of "I realised I could not officially be an old man unless I was driving a Rover". It's up to you to decide whether I was completely making shit up.)

This car spent most of its life in South Africa. The chap that owned her brought it over here about four years ago, then got ill, and died. Penny made her way to a dealer, who then advertised it on Car and Classic. There was a "Project Profile" feature piece, which I read, and I immediately knew I had to have that car.

I've had a bit of a soft spot for the P5 since I saw one at the end of Lock, Stock and Two Smoking Barrels, but I do not think I have ever been emotionally affected by one. I don't think any car ever has affected me as much emotionally as this one did via mere pictures on the Internet.

I sat on those "I need this car" thoughts for an entire work day. When I clocked off, I called the chap and provisionally reserved it. Exactly one week later, it was delivered to my driveway. And she is every bit as beautiful as she was in the photos.

So what's the plan?

My immediate plan, as in tomorrow lunchtime and this weekend, is to make everything watertight. Brand new door seals are included with the car, but not fitted. I'd also like to make the boot latch work (it does not work), and the bonnet to close (even better if I can open it again).

Next up, of course, I want to make it solid underneath. Because the car spent most of its life in South Africa, I am hoping that I will not have too much to do. South African weather is much kinder to cars than our own climate is. But the car is 57 years old, so no doubt I will be able to poke some holes in it.

In the meantime, I will have to go through some paperwork and some small amount of arse to get it registered in the UK. The guy who imported it never did. I am hoping this will not be too difficult. While this happens, I will not make any modifications to it - and by that, I mean not even replacing any parts with functionally identical modern ones. My reading of the rules for getting an age-related plate when "naturalising" a car is that every component must be original. My reading might be unreasonably strict, but I'm erring on the safe side because I do not want to end up with a Q plate.

After that, will be the slow process of recommissioning; replacing every consumable part, checking every safety-critical part, getting the engine working (it may or may not work, though it does turn over by hand), source the very few bits that didn't come with the car, and eventually getting it to a state in which I can MOT it. (I'm quite aware that at its age I don't have to MOT it, but with the amount of money this will cost me it would seem foolish to not spend an extra few dozen quid for an extra set of eyes to ensure that I have not done something stupid). My hilariously-over-optimistic timeline for this is to have this done before the end of Summer.

After that, let's talk about what I will not do. I want to do nothing that will change the car's character. It was designed to be a big, boaty cruiser for important people, and it will continue to be a big, boaty cruiser for some idiot from Norfolk. It is a silent straight 6, so it will stay a silent straight 6. (And if I was going to do a 3.5 litre V8 swap I've got one kicking around which is much more interesting to me than a Rover one.)

I also don't want to change the car's history. There are dents and scars on the body, and those are part of the car's past. I don't think I should attempt to do much with them. I probably will stabilise those scars where they are showing surface rust, so they don't turn into something worse. And even though the metallic paint is almost certainly not its original colour, that too is part of the car's history and not something I want to change. Plus, I love that colour.

What I will do are some tiny modernising touches. I'll definitely convert it to use an alternator & to have negative earth electrics, and replace any invisible electronic components with modern solid-state ones. It will definitely get seatbelts at some point, because I do not want the tiniest low-speed shunt to cause a brief, very exciting trip through the windscreen. I might do something about the distributor having points, and about the fact that viscous fans are wank...and probably a lot else, but you get the idea. Tiny modernising touches, as I said, that will make her a more viable daily.

That's all for now. Onwards!