N52 Swap Discussion

[Bored]

To add some more substance to the equation:

https://www.e90post.com/forums/showp...4&postcount=36

That link has a download for a pdf looking at the technical aspects of the N52.

At 8,000 RPM, the mean piston speed of the N52 (if the stroke is 85mm) is 4,462 feet per minute/22.67 meters per second. The Honda S2000's F20C at 9,000 RPM (84mm stroke) is 4,961 feet per minute/24.6 meters per second.

I bring up the S2000 as it is a factory 9,000 rpm car with all the warranties that go with it. The 4961 mean piston speed is much higher than the 4462 than the N52 would see at 8000 rpm. We could even continue to push the redline higher, but somewhere between 8200-8400 its been noted that the hydraulic lifters start failing.

The N52's smaller bore than the S2000 likely reduces total piston-wristpin weight below the S2000's. Weight optimization in the N52 likely leads to a lighter connecting rod even if its longer, as it has less mass on the end of it.

All this leads me to believe that if we can solve the hydraulic lifters pumping down at high RPMs, we can reliably spin the motor higher than 8000 rpms for extended periods.

This leads to other issues to be addressed before we can add more RPMs and make power.

Harmonics, the big enemy of the straight six. Thankfully, ATI already has an N54/N52 crank damper that fits our application. I was hoping that Fluidampr would have a product for this application but it seems not. This doesn't SOLVE the harmonics but simply mitigates their effects. I don't have ANY idea where the harmonic resonance is on this crank, and crankshaft whip is likely part of that...

The three stage intake manifold stifles power past 6300 as demonstrated on many n52 cars dynographs. The N54 intake manifold does better, especially if the manifold is shifted to better align with the ports in the cylinder head. A custom intake manifold could be drawn up to accommodate high RPM usage.

Cylinder head flow is another POTENTIAL bottleneck. I haven't found any flowbench information on the N52 head. All pictures of the ports show a high quality casting which on the texture level is unlikely to be improved for port injection. Valvetronic complicates valve seat/angle jobs, to what degree I'm not sure. Research and development here will lead to gains, but how much is hard to state.

Camshafts... the biggest way of changing the character of the engine. RiotRacing had some issues with their attempt at N52 cams, the regrinds failing to be useful. I've heard talk of a potential exhaust camshaft development being looked into (to compliment MILVs), but nothing hard nor fast. The exhaust cam set-up is much simpler, potentially leading to easier/cheaper gains there. Intake cam development is more difficult due to valvetronic being used to maintain idle.

Oiling has already popped up but I'll touch on it again. Hydraulic lifter bleed down in the valvetrain is likely the first order that needs to be corrected if a radically increased redline is to be sought. Due to the variable/displacement on demand on the oil pump, I cannot be sure simply porting the pump will increase oil flow... nor have I found any numbers citing potential cavitation at rpm of the pump. If the bleed-down of the lifters is oil pressure, then there might be a way to add a restriction oil flow to the lifters themselves to increase pressure... but this is simply theoretical conjecture.

Hopefully all of this stimulates some conversation.

[hubcapboy]

I’m mostly just trying to get the engine into the car.


Sent from my iPhone using Tapatalk

[The Dark Side of Will]

Just design it to work when annealed. It's not difficult to make something strong as the loads aren't high. I'd be worried about weld fatigue cracks over time which rehardening doesn't really address.

One could, but I haven't seen it done in the BMW community.

It's also possible to 3D scan the engine compartment and do an engine swap in CAD before every trying the real thing. The 3D scan is probably more likely, as FEA/FEM requires some skill that's rare in the aftermarket. We also don't know what the loads really are for engine mount arms. We can guess with free-body diagrams and apply rules of thumb for dynamic loading, but those would only be estimates.

You could also bolt a tapped alloy pad to the block and bolt a steel fabricated bracket to it

Totes. It's a thing.

[Motheye99]

On the topic of better airflow, I've been working through trying this on a M54 using a M50 intake. But i suppose it could be applied here. I like the engineering happening in this thread

My plan is to cut the intake leaving the flange for the head and runners of what ever length your motor needs. The blue hash represents the removed part of the intake.


Because these manifolds are plastic to begin with I believe a properly 3D printed adapter could be epoxied to the cut side of the N52 intake to allow adaption of s54 TB's. Nylon Filament prints at 450f+ So with shrouding the adapter hopefully wouldn't warp. CNC delrin or aluminum could work as well.

It wouldn't be hard at all to duplicate this S54 intake flange design on a adapter.

[hoveringuy]

Does the N52 even need Throttle Bodies??

[hubcapboy]

I think somewhere in a Bimmer Performance Center video they make reference to having successfully run the engine without a manifold mounted... just the ports on the head sucking air.


Sent from my iPhone using Tapatalk

[nando]

No it doesn't. However, as I've discussed many times, without any throttle at all - if you have a valvetronic failure, the car would either be dead or would suddenly go WOT.

Also, whats pictured there is an N54, not an N52 - the ports on the head are much shorter than the N52, which has the injector holes built into the head. That is why I think the S54 throttles should work well on an N52.

I test fitted it two years ago:



A very slightly tapered adapter would be needed. MSV70 can control the S54 thottle actuator, but some minor modifications may be needed (it uses more current, so you'd want to beef up the traces to the driver and use a different pin to connect it to the harness).

The biggest issue is the first throttle interferes a bit with the OFH. I think it's solvable, but you'd have to modify the throttle flange or the OFH bolts or both.


The injector ports on the S54 throttles would be blocked off. You'd have to make an adapter to hook up the N52 MAP sensor to the vac tube on the throttles as well (similar to the CSL). And of course you have to bolt some sort of throttle actuator bracket to the block.. which can't be made of steel.

The point of all of this isn't having ITBs, but to have something that will work up to 8000rpm. As was mentioned before, the N52 3 stage manifold chokes the engine at high RPM (the torque drop off is actually 6900rpm, not 6300rpm - the N52 peaks at 6700rpm). The N54 manifold "fits" but is hardly even close to ideal - it's just cheap and happens to bolt up (even if the ports don't line up at all without modification).

Regarding cams - yes, I believe Marty was working on an exhaust cam. I have no idea what progress he's made. It's basically a regular cam so it's definitely doable. Although without headflow data, who knows what it really needs.

ForcedFirebird has had an N52 head to flow for a long time, but unfortunately, Pete, who was bankrolling that project, tragically died in a motorcycle accident. I think he still wants to flowbench it anyway but it would take funding.

I don't see any reason why it couldn't make 300hp at 8k though. So far nobody has gotten close to that because they're using N52 or N54 intakes and basic bolt ons.

I honestly don't think we need to worry about the oil pump too much. It's basically designed so it can flow way more than it needs to, but not too much at any given load. Bleeding out of the lifters IMO is best solved with solid lifters and shims - like the S54. Keeping in mind the *stock* redline is 7000rpm - pushing that 10% higher isn't likely to require significant modifications to structural elements.

[Motheye99]

Also, whats pictured there is an N54, not an N52 - the ports on the head are much shorter than the N52, which has the injector holes built into the head. That is why I think the S54 throttles should work well on an N52.

I made this quickly in solid works, Just to give a rough idea to what I mean. The reason I like this design is it allows me to use 3D printers to mate the S54 ITB's to the N5X head. This is would fix the imperfect geometry the s54 ITB's have when directly mate to the head.

And correct me if im wrong but it is good to have some runners for a motor to help with torque. Weather it be before or after the throttle body. The e46 M3's air box has pretty long velocity stacks.

[nando]

Good idea, but you'll want to make it much shorter and I don't think it need to angle, or you'll never fit an airbox.

We already have some runners. The N52 head has part of what was traditionally part of the intake manifold built into it. Once you figure in an adapter 1-2" thick, that should be plenty. The S54 is much shorter, remember we're targeting a power band that peaks near 8k.

Also you'd still have to correct the geometry anyway, because the N52 runners are smaller than the ports at the head. It would be better to ditch the N52 runners and just make your own adapter flange. Could still 3d print it.

I made a template in CAD a while back, I can post it here.

[Motheye99]

Good idea, but you'll want to make it much shorter and I don't think it need to angle, or you'll never fit an airbox.



We already have some runners. The N52 head has part of what was traditionally part of the intake manifold built into it. Once you figure in an adapter 1-2" thick, that should be plenty. The S54 is much shorter, remember we're targeting a power band that peaks near 8k.



Also you'd still have to correct the geometry anyway, because the N52 runners are smaller than the ports at the head. It would be better to ditch the N52 runners and just make your own adapter flange. Could still 3d print it.



I made a template in CAD a while back, I can post it here.

Ok very interesting. I have no knowledge of the n52. However I plan to get one next sale at my local pyp. Please post that adapter I think I could get some use out of it.

My main concern with 3d printing a adapter is the prolonged temps it would directly receive from the head. However the oem intake is also plastic.

I originally intended this for a m5x motor hence the longer runners.

[wazzu70]

You can 3d print a manifold, you just have to use the proper material.

A while back most race team prototypes were made with Windform material which uses the SLS process instead of the more common FDM.

The key is the material, you can’t use a hobby type “ABS”

[digger]

It's still not very economical to 3D print a manifold that will actually last long term . Some of the less critical areas of the manifold would work fine. Bolting to the head no way. A Cnc or cast adapter is the way to go IMO depending on quantity

[Bored]

Still looking at this swap as a viable option as I just got my first E30. '91 kaschmirbeige 325 auto coupe. Wasn't tickled about the automatic, but the car is in really nice shape otherwise...

I'm loathe to take the car apart at this point, plus I live in an apartment complex so there is nowhere to really tinker with things or store 'em.

Still running ideas through my head on how to fit this to work. One of my neighbors works at an auto shop (race cars, custom cars, etc) and he has some skills in fabrication... might see what he can do with it.

[nando]

With no space to work on it I'm not sure how you would make it work. Paying a shop to do this swap would cost an astronomical amount of money, unfortunately.

The primary fabrication is the oil pan/subframe and mount arms. Everything else for the most part is like any other swap.

[The Dark Side of Will]

A very slightly tapered adapter would be needed.

Why an adapter? It bolts up, why not port out the N52 head to match the throttles? The ports and, I @$$ume, the valves are smaller than S54's, but we don't yet know if the N52 head can be improved to perform closer to an S54 head. I suspect that someone looking to swap on S54 throttles could afford a head/throttle package.