[DRIVING] Audi A2 /w Prius Gen2 Transaxle & Inverter [FINISHED]

[arber333]

That didn't work too well... The handbrake is all crusty and won't release and was stronger than the glue. Car was still on ramps. I knew what would happen - at some point the ramps just shot forward. Just didn't expect them to puncture my Loctite cleaner bottle

No gotta get the whole thing out, clean it properly and glue again. If it don't hold up get a driveshaft made

As it is not shrink fit i would advise against glueing. I think your best bet would be to bevel the larger shaft edge on a lath and drill two 8mm holes into larger shaft spread some 10cm apart 90deg shifted relative to each other.
Then find a good welder who can make a circumference weld first and then fill both spot welds.
Then you would need to put your shaft to lathe to balance it; or you can fit a rubber damper to it... they used to do that in the past for dynamic balancing and it works somewhat . I think rear Outlander shafts come with rubber dampers...

You could also temper the shaft at 400deg or similar for some time to relieve straess of welding... technically correct term is normalizing . But that would be worth if you would fit your shafts to some high power conversion...

[johu]

I think you're right but I'll rather have a shaft made by motomax.pl instead of going through the trouble you described above

[Ev8]

If only you were local I’d come and weld that up for you so you could get testing!

[RetroZero]

Me too .

[johu]

Todays result: driveshaft will go to Poland

I spent many hours renewing the rear drum brakes. Had to cut the one drum apart because it wouldn't come off.

With that done I drove the car up the driveway. Parameters still need a lot of tuning, it drove quite jerky with a tendency to over current trips. At some point some transient broke the glued sleeve again, so now I will have it made professionally

[RetroZero]

I got a local turner to make a sleeve between the solid and hollow shaft, and then went to another company to get it welded up. No one around where you live? I told them what I was building and they were all ready to help. I even mentioned them in local news, which helped too... Just a thought for a quick turnaround.

[Ev8]

Well congratulations on it moving even if it was only briefly

[johu]

Thanks, yes actually it was the first electric movement

Re welding: yes I could probably run around and have that done but I'll rather have it done properly and Thomas didn't object to the extra cost. He did object to welding though because he has seen so many broken welds in his life

[Ctwidle]

Before i had my beetle gearbox rebuilt i was considering having half shafts built to join the beetle hubs to the Outlander diff. I approached a local engineer who specialises in preparing high powered race cars. He was not keen on an annular welding between a sleeve and a shaft due to the need for normalising the heat induced stresses. His proposal, which i haven’t seen on here before, was to cut the sleeve at angles to increase the length of the weld and spread the area of stress. I hope my attempt at PAD (paper aided design!) demonstrates the technique.
Chris

[eee291]

viewtopic.php?p=29272#p29272
It has been done before, this requires welding and machining, though.
I imagine just about everybody has a welder or someone local to them. Machining/turning the sleeve can be a bit difficult.

[lsh3rd]

In a similar vein to what Ctwidle posted, I think this approach is a bit more DIY friendly and nearly as good:


(Sorry, language in video is a little crude, but the guy probably knows his stuff.)

The approach probably wouldn't work well for the hollow VW shafts in Johannes's case, but for general use, I think the approach looks sound.

[arber333]

Before i had my beetle gearbox rebuilt i was considering having half shafts built to join the beetle hubs to the Outlander diff. I approached a local engineer who specialises in preparing high powered race cars. He was not keen on an annular welding between a sleeve and a shaft due to the need for normalising the heat induced stresses. His proposal, which i haven’t seen on here before, was to cut the sleeve at angles to increase the length of the weld and spread the area of stress. I hope my attempt at PAD (paper aided design!) demonstrates the technique.
ChrisCA027CE1-2F48-40A3-BD53-F7EE7D5528D0.jpeg

Yes this technique works similarly to my previously described welded holes after circumfverence weld. As torque acts in 90deg relation to force applied you are extending your effective handle on the shaft and thus material sees less peak torque at the joint. Its all about torque distribution.
You could also make a V shaped cut in the larger shaft on both sides which would effectively act the same...

[johu]

Finally a big update here.
Thomas came over to Kassel and we worked away on the Audi. Properly mounting 12V battery and various other housekeeping.

We also test fitted the A/C compressor

To our surprise yesterday the drive shaft from motomax showed up

Of course we put it right in and then assembled the car again

Then it was time for a test drive. It doesn't like low speed and low throttle but once at speed it drives very smoothly. We were going to go onto an extensive drive when we remembered there is no oil in the gearbox. So drove back to the driveway and filled it with some generic transmission oil. After the 3rd bottle it started leaking out from the left shaft seal
And today Damien tells me that it needs special "WS" oil otherwise the enamel of the stator is eaten over time.

Anyway, very happy with the results and very impressed with motomax. More extensive test drives have to wait until the oil and seal woes are fixed

[Zapatero]

Nice driveshafts, looks familiar to four that i have bought from Motomax.pl

Looking forward to the first real test drive video and i'm very curious about power, range etc..

[RetroZero]

Good to know it's rolling

[johu]

Test drive

[janosch]

... Had to cut the one drum apart because it wouldn't come off...

Oh god, our very early prototype had a drum that was badly seized:
pulling technique on YouTube <-- this is what I did back then. The twisting big bolt in the middle there didn't work for me, I used a car jack to rip it the drum off with an improvised grabber similar to what he has in the video there.

I am sure "real" mechanics have endless stories about this stuff.

[johu]

More test driving today. Shaft seal fitted, driveshaft cup now fully in and elongated. I noticed it is far from straight as the transmission sits rather far forward. I noticed normally they should be straight. Is it an issue?

Anyway, the test drive. The car accelerates fine to 40 kph then looses torque. With FW disabled it will go up to 50 kph, that is pretty much it. With FW (both variants) the PI saturates and the car accelerates on its own. It seems that uq would need to rise (or become more negative, we are spinning in reverse) in order to lower the current. Strange things.

Anyway, the inverter takes it like a star. I had to force-stop with the brake meaning the current would rise the slower I went. iq peaked out at 420A (plus some id) on a Prius inverter! It did not trip. It trips in unsuspecting situations during regen or low throttle sometimes but restarts without issues.

Alright a lot to do but because voltage runs out at 40 kph it's a lot less dangerous than with Touran. Video coming.

[Pete9008]

Sounds a little like +Id might be being added rather than -Id. Is it possible with spinning in reverse that a sign has got reversed somewhere?

[johu]

hmm, good question. Don't see any dependency on direction in the id data path.

Will try today with your syncadv mod plus a manualifw parameter that just lets me dump some fixed amount of id. That way I can also check whether that creates unwanted torque

[Pete9008]

What speed is the motor turning? If it's fairly low then I wouldn't expect the sync mod to do much but equally it shouldn't do any harm.

The really odd bit is it running faster with FW disabled and the PI saturating with it on. They both make me wonder about Id polarity. Manual Id is a good idea and should allow you to tell.

Edit - this could be complete rubbish (as I can't quite get my thoughts straight) but could tuning to the wrong syncofs null essentially reverse Id? Or alternatively can PM motors be built mirror image so needing a sign reversal?

[Ev8]

Exactly what I seem to encounter with f/w enabled on my gen2 inverter

[Pete9008]

Exactly what I seem to encounter with f/w enabled on my gen2 inverter

Did you see it run to a higher speed with FW disabled too?

[Ev8]

To be honest as soon as I’m in f/w region and it accelerates on its own I don’t let it keep going to so what it tops out at!

[Pete9008]

Good point!

Hopefully Johannes tests will find something that explains what's going on.