3.2 Turbo Build

[greglepore]

That's crazy nice. You could probs offset your cost by selling the 3d plans for that...

[Coultl]

Thank you! It’s mostly CAD and plastic mock ups at this point. Well so how it all actually comes together!

[Coultl]

I posted this in my engine rebuild thread but since it’s directly related to the turbo system I’ll post here too.

To lower the compression ratio I took some material off the top of the pistons. I ended up pulling exactly 10 grams off each piston. I would have liked to take a little more off but it felt like I should quit while I was ahead. To get those 10 grams I pulled about 6mm off the top of the piston.

With Aluminum’s density of 2.75 grams per cc, the 10 grams I pulled off is equivalent to 3.6cc of aluminum…removed from the combustion chamber. This lowers the compression ratio from 9.5:1 to 9:1.

I think 9:1 will be passible for my modest power goals, but it a still a little high. I ended up going with 1.3 mm base shims on the cylinders. My stock base shim is 0.5 mm because the case halves have been machined. The additional 0.8 mm increases the combustion chamber volume by 5.7 cc. Adding it up, the machine pistons plus the base gaskets resulted in a compression ratio of 8.4:1. Perfect.

I have not seen anyone else cut these pistons down for this purpose but there’s a lot of aluminum up top. I think they will be fine.

[Bucketlist]

Your a brave man or maybe just smarter than the rest of us!

[WonkoTheSane]

Glad to see someone else experimenting with turning their pistons down. I've been curious about doing that myself too when I can eventually get into my engine.

[LukasM]

Just discovered this thread after following along your rust repair one Coultl, which I enjoyed a lot.

Very cool work on the 3D scanning, design and printing! And it will be immensely satisfying to eventually drive a car that you rebuilt from front to back yourself.

One thing I would caution on is using FDM for structural parts that need isotropic strength, so as a pressurised intake manifold. Even with a 3mm wall thickness and a strong material like PPA-CF, the layer adhesion is most likely going to be an issue. The added CF works well for helping appearance, stiffness, and preventing warping, but it also weakens layer adhesion since it also acts kind of as a contaminant where the plastics wants to stick to itself. The newer core filaments such as from Siraya Tech partially address this issue, but strength in Z is still much lower than along the layer.

I wonder how much SLS printing this big piece in China would cost, either in nylon or ideally some form of metal? Alternatively, you can also try to find somebody with a printer with a heated chamber and an annealing oven, both of those can also help the parts become more isotropic.

I would definitely build a testing rig you can pressurise to play around with it, before it goes on a very expensive Porsche motor. Please keep us posted on how this testing goes, hopefully I am completely wrong and it all works out sweetly in the end.

Cheers,
Lukas

[Coultl]

I hear you on the strength issues. Been thinking about it.

I’ve actually been using the core PPA from Siraya. It prints great. Later adhesion is better but still far from the material’s actual strength.

The one thing that I really don’t like is that a failure of the intake manifold could create a large vacuum leak which would be a lot like a stuck throttle. A small crack in the intake plumbing would be just be an inconvenience but a large crack after the throttle body is a safety issue.

It may be that I prototype with 3D printed PPA and then end up having it 3D printed in aluminum once it’s all sorted out. We’ll see.

I’ve been comparing my parts to the BMW M3 intercooler end tanks I have. They are pretty stout glass filled nylon but people run well over 20 psi boost on those cars. Pretty amazing what the plastic can handle. I’ll have roughly twice the wall thickness.

[LukasM]

Good to see that you are also into the details of 3D printing, it really is a cool technology for us tinkerers and making advances all the time.

The big difference with those BMW end tanks is that they are most likely injection moulded, and therefore as strong in every direction. I don't recall Siraya Core off the top of my head but most PPA-CF's have about 1/3 of the strength in Z between the layers, as in X and Y.

I'd probably trust it myself with a NA engine if printed in a heated chamber, but boosted is a whole other ballgame...

Have you checked with your CAD drawing how much an alu print would cost? Probably not cheap even in China but we all know how expensive it can get to rebuild an air cooled engine...

[Coultl]

Yes, the BMW end tanks are far stronger for a given material thickness. They are also more heat resistant than the PPA. The PPA should handle heat better than PA-6 but 3D print filaments have modifiers and such. I tested and it gets soft first.

I print in a heater chamber so it will be as strong as PPA can be…but for a given material thickness, it’s still far from an injection molded part.

That’s said, my design is absurdly overbuilt. All the math will likely show a factor of safely of 5, maybe closer to 10…but real world is a different ball game…

I’d need to redesign it for aluminim and it would kinda ruin the look….but it would be doable. I think total cost would be about $1k.

[flightlead404]

Originally Posted by Coultl

I posted this in my engine rebuild thread but since it’s directly related to the turbo system I’ll post here too.

To lower the compression ratio I took some material off the top of the pistons. I ended up pulling exactly 10 grams off each piston. I would have liked to take a little more off but it felt like I should quit while I was ahead. To get those 10 grams I pulled about 6mm off the top of the piston.

With Aluminum’s density of 2.75 grams per cc, the 10 grams I pulled off is equivalent to 3.6cc of aluminum…removed from the combustion chamber. This lowers the compression ratio from 9.5:1 to 9:1.

I think 9:1 will be passible for my modest power goals, but it a still a little high. I ended up going with 1.3 mm base shims on the cylinders. My stock base shim is 0.5 mm because the case halves have been machined. The additional 0.8 mm increases the combustion chamber volume by 5.7 cc. Adding it up, the machine pistons plus the base gaskets resulted in a compression ratio of 8.4:1. Perfect.

I have not seen anyone else cut these pistons down for this purpose but there’s a lot of aluminum up top. I think they will be fine.

I'm curious about the shape left after machining. IDK if the flat top left impacts "swirl" and charge mixing, but the sharp edges worry me a little. I guess time will tell.

[Coultl]

I smoothed out the sharp edges somewhat afterwards. Valve pockets typically have hard edges and that’s a sign that it might not be an issue. We’ll see!

[Coultl]

The design of the turbo mount took a lot of work. The solution wasn’t obvious but I’m really happy with where it landed.

The mount bolts to the turbo with two M14 bolts threaded into the turbos coolant connections. These are big threaded connections to the core of the turbo. The mount connects to the engine via four of the valve covers bolts. These are M8 and thread into chunky aluminum on the cam house. I don’t love that the load is going through the cam cover gasket, but I don’t see any other option.

The thing that’s got me excited is that the oil drain is now integrated into the mount. The oil drain is sealed to the mount via an ID/OD o-ring seal. The mount is then sealed to the cam cover via the same rubber compression ring used for the oil pump / case connection. I only need to make a corresponding hole in the rocker cover. It’s all very compact. That said, there’s no drop and the oil has to travel across to the valve cover. Not ideal but these ball bearing turbos use so little oil (1mm restrictor) that I think it will be fine.

[Coultl]

Photos of the prototype. Feel sturdy even in plastic but it will be 3D printed aluminum soon. Installation is cramped but not too bad. Adjusting valves will be quite the job however…I’m guessing that’s why Porsche didn’t put the turbo in this location.

[moparrob]

Wow, you have made alot of progress and that is looking good. I can't wait to see the results with the parts printed in aluminum.

[reclino]

Wow, this is an impressive build, very fun to watch and nice to see some data driven decisions being made. What scanner are you using for this project?
Thanks
David

[Coultl]

Additional photos below with some key plumbing in place. Seeing it in person usually results in some issues and opportunities to show up…but nothing yet. Still feels good.

The very tight fit for the compressor exit hose is unfortunate but this was clear and CAD and moving the turbo forward isn't an option (the wheel is in the way) and moving the turbo aft would require major frame cutting. Pulling the coil off will be annoying but it works. I'd say it's still 50% easier than working on a modern car...

Insulating the turbine housing is something I’m thinking about. It’s going to be quite close to the chassis and engine tin. I’d rather not use a turbo blanket.

[Coultl]

The metal 3D prints arrived today. $200 delivered + $60 tariff… Not bad.

The mount is aluminum and the oil drain is stainless. The quality is VERY good and there’s no detectable warping.

The turbo feels so solidly mounted. Love it. Next up are the stainless exhaust 3D prints.

[quattrorunner]

That is fricken genius!

[moparrob]

Those parts are beautiful. Which vendor did you choose for the powder printing?

[Coultl]

I use CraftCloud and the vendor is use is 3DNext. A little higher cost but seems like the quality is higher.