Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Topics concerning the Tesla front and rear drive unit drop-in board
User avatar
asavage
Posts: 328
Joined: Sat May 14, 2022 10:57 pm
Location: Oak Harbor, Washington, USA
Has thanked: 269 times
Been thanked: 103 times
Contact:

Rotor bearing at gearbox end requires locknut removal to replace

Post by asavage »

The rotor bearing at the gearbox end has a groove nut. It's the equivalent of SKF KM 7.
SKF (and others) KM 7 axial nut ("groove nut")
SKF (and others) KM 7 axial nut ("groove nut")
(https://smile.amazon.com/Whittet-Higgin ... B01EMOZQNU)

image.png
(https://youtu.be/_GigD0s47RI?t=129)



While R&R of this nut doesn't necessarily require the mating tool (workarounds exist, including a nice SKF Hook Spanner wrench), the SKF TMFS 7 socket is a "correct" tool, and allows torquing the nut to a specified value if desired. I ordered one from Acorn Industrial (UK) for USD$88 + shipping (USD$112 shipped to my doorstep, six days delivery via DHL).

SKF KM 7 Axial Nut Tool
SKF KM 7 Axial Nut Tool
SKF KM 7 Axial Nut Tool
SKF KM 7 Axial Nut Tool
SKF KM 7 Axial Nut Tool
SKF KM 7 Axial Nut Tool
Al Savage
2014 RAV4 EV
NissanDiesel
WimV
Posts: 25
Joined: Sat Mar 12, 2022 10:21 pm
Been thanked: 3 times

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by WimV »

In attachment some pictures of my triple lip rev Q seal with a small leak for at least 8 months. The seal is 5 years old and 77.000 km.
All the seals are flat without concentric rings or other features. The seal at coolant side and dust lip were in pretty good shape. The middle seal was almost completely gone. There was a black sticky crust of dust between the two coolant lips, this was probably the disintegrated PTFE from the middle lip.
My rotor shaft surface was in pretty bad shape.
Attachments
IMG_20221021_165939.jpg
IMG_20221021_165729.jpg
IMG_20221021_165349.jpg
IMG_20221021_165242.jpg
IMG_20221021_165044.jpg
IMG_0004.JPG
IMG_0003.JPG
IMG_0002.JPG
IMG_0001.JPG
WimV
Posts: 25
Joined: Sat Mar 12, 2022 10:21 pm
Been thanked: 3 times

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by WimV »

By the way: I’ve checked the speed sensor after driving 400 km with the new seal (Ceimin CTLBDYW). Well, that wasn’t what I hoped. There was condensation on the sensor. The drain was still dry (was plugged, so I would know how much it leaked). I collected about one drop of coolant when using some compressed air to flush the reluctor chamber.

I don’t think the speedi sleeves are usable with PTFE seals. I think the seal is stretched too much cause of the increase in diameter.

I’m going to keep driving for a while and keep checking the leak. If it get worse, I’ll pull the rotor to get it welded and grinded to the correct diameter and try with a new seal.
Attachments
IMG_20221014_114103.jpg
IMG_20221014_113832.jpg
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

WimV wrote: Sat Oct 22, 2022 7:49 am By the way: I’ve checked the speed sensor after driving 400 km with the new seal (Ceimin CTLBDYW). Well, that wasn’t what I hoped. There was condensation on the sensor. The drain was still dry (was plugged, so I would know how much it leaked). I collected about one drop of coolant when using some compressed air to flush the reluctor chamber.

I don’t think the speedi sleeves are usable with PTFE seals. I think the seal is stretched too much cause of the increase in diameter.

I’m going to keep driving for a while and keep checking the leak. If it get worse, I’ll pull the rotor to get it welded and grinded to the correct diameter and try with a new seal.
PTFE seal studies says leak is unavoidable and quote 1mL/hr @ 10m/s surface speed for 30w oil. So I expect we likely will see leaks fairly quickly with any seal. Its interesting speed sensor shows leakage sooner than drain which make sense. Most drain mod monitoring is probably via the far easier drain tube. Don't really understand how speed sensor can be dry after some miles.

Read the following leak control solution in a seal book. The rotor/stator in diagram doesn't refer to LDU but the slinger and beveled collection structure provides some insight to what happens to the leaked fluid. I thought maybe the speedi sleeve lip might act like a slinger but we lack the bevel to concentrate the leak for removal. Speed sensor chamber probably act like a trap for leaks.
Screen Shot 2022-10-22 at 3.12.04 AM.png
Screen Shot 2022-10-22 at 3.21.19 AM.png
Thought about making a borescope access hole to peek inside reluctor chamber. Just not much room and no space to go through speed sensor opening with reluctor wheel blocking access.
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
smathermather
Posts: 12
Joined: Fri Dec 02, 2022 5:35 am
Has thanked: 7 times

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by smathermather »

Hi All, this is quite the thread and breakdown on the LDU.

I'm recent 2017 MB bclass owner (b250e), and would you know it: it failed in the way one might predict, 3 or so weeks into owning it. Fortunately, while the seller isn't my favorite car vendor these days, they did provide an extended warranty under a fantastic provider who surprisingly paid for a replacement motor rather than totaling out the vehicle.

The reason I write is, well first to say "Hi" and thank you all for the amount I've already learned on the forum, and second: they aren't putting the original motor type in as it is no longer available (probably all the failures...). Instead they are putting in a "redesigned" motor, which also requires new programming which we are waiting on from MB HQ. Now, I don't know what this means yet, whether this is just an LDU with a better seal or some other motor design, but as I find out more, I'm happy to report on what I can. If it is a redesign of the existing motor it could be interesting to this discussion.
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

My LDU rebuild has been done for a few weeks now. Used Ceimin yellow triple lip on 600->1500 grit prepped shaft surface. 400mi and no leak on speed sensor so far.

Am still chasing quality brand PTFE seal source for the community. Sharing a status update. Prior update here.

https://openinverter.org/forum/viewtopi ... 858#p46858

Quick summary is no new source but can offer a Parker custom seal PN that others can also chase down if interested

Water Pump Style Ring Faced Seal

Found a company https://www.ergoseal.com/ that Tesla tier suppliers contacted for seal design a couple of years ago. They showed a drawing of a water pump style seal but never got beyond drawing proposal phase. So DIY community can't afford the dev+test cost.

I'm no water pump seal expert (nor a PTFE/FKM seal expert either haha) but this solution seems quite challenging to me. Regular water pump seal is factory machine assembled into the water pump along with leak drain chambers/weep drain channels AND an additional barrier of some sort between leak chamber and bearing.

https://www.agcoauto.com/content/news/p2_articleid/190

LDU's structure seems to be quite challenging for this from 2 areas

- Space is ultra tight for a water pump type seal's structure and its internal components along the axial direction
- Likely need to assembly the seal by mating its 2 halves so each halve can be water tight pressed onto their mounting structure (coolant shaft, coolant manifold) Any high precision assembly tolerance is not achievable with the way coolant manifold mount onto the shaft.

Regardless, the big barrier is can't pay for custom dev+test.

Triple Lip Seal

Struck out at Saint-Gobain's custom seal division on prior update. Sounds like they are bound up by Tesla in some way. Reached out to Parker-Hannifin's similar dept (Engineered Polymer Systems - EPS). Talked to 2x application engineer, 2x customer support ladies. 1x local distributor. Dead end so far but sharing a PN and spec from Parker's custom catalog

EPS catalog 5340 (PTFE seal) Dated 2008-2018 https://www.parker.com/literature/Engin ... 0_2008.pdf
EPS catalog 5350 (Rotary seal, inclusive of PTFE) Dated 2006 https://www.parker.com/literature/Engin ... s/5350.pdf

Both catalog allows choosing from many parameters for desired spec. Many overlap info between the 2 catalog but also complementary info in PTFE material property tables.

Provide the following seal PN to distributor to quote below. They've not replied after a couple of weeks so will need to follow up.

M 0615 LDE 1250 03000 080 N

Untangling this spec is found on page 8-2 of 5340

M=metric

0615 = Proprietary low wear PTFE material
8-6 3-2 3-4 3-5 8-6 of 5340
9-8 of 5350 include min shaft hardness (we can't determine without testing so just guess something softer rather than harder 25Rc for 0615)

LDE = LD (dual lip) + E (excluder)

1250 = 12.5mm which is half of 55mm (OD) - 30mm (ID)

03000 = 30mm ID

080 = 8mm width (don't know how far lips will extend beyond seal width (Ceimin stretches to nearly 12mm leaving 1mm+ of margin left on the shaft to margin both coolant and excluder lip) see pdf doc here ( https://teslamotorsclub.com/tmc/posts/7185566/ )

N = NBR for o-ring. Can inquire about Parker's special coolant resistant FKM later which surely requires more Parker engineering discussion. https://www.parker.com/Literature/O-Rin ... G_5040.pdf

Calling Parker EPS (# on catalog) can not reach app engineering directly but support ladies will route. 2 application engineers I spoke to wasn't too helpful. They mostly confirmed what I already knew and one said just go with OEM thinking OEM has done their homework. They confirmed the following

- Lathe cut depending on material and spec. Diameter of each lips are automatically determined by ID spec (and maybe material?) Ceimin seals is probably like this. Primary coolant seal lip is tightest at 26.5mm unstretched and excluder lip at 28mm unstretched ( https://teslamotorsclub.com/tmc/posts/7185566/ ))

- No pump aid structure (same as Ceimin and Tesla single lip PTFE OEM)

- No concentric rings to aid bending (same as Ceimin) Tesla single lip PTFE OEM has rings

- Caged seals are discontinued (shown on 5340 catalog) so have to go with o-ringed on OD and probably can sealant mount into bore.

====

Also want to note the triple lip PTFE seal installation is a highly error prone step. Almost impossible to not fold the excluder lip unless you know the assembly tricks. The same link provided above outline this in full detail

https://teslamotorsclub.com/tmc/posts/7185566/
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
User avatar
Boxster EV
Posts: 434
Joined: Fri Jul 26, 2019 9:32 pm
Location: UK
Has thanked: 35 times
Been thanked: 26 times

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by Boxster EV »

Data point and update: after 3500 miles my SKF seal is leaking and progressively getting worse (I have a drain plug so it's dumping coolant overboard). I can see the expansion tank level dropping every few hundred miles and am having to keep it topped up.

I'm not going to have time to remove and examine it before Christmas, but I thought I'd post an update before someone follows my lead with the SKF route (IE don't). I'll take pictures once out of the car.

As a reminder, the SKF seal I used is actually designed for oil applications so it was a long shot expecting it to last any reasonable length of time.

Looks like I'll have to get my rotor shaft repaired and refinished (so that I wont need a speedi sleeve), however we're all still stuck on finding a decent replacement seal.

Edit 20th Jan 2023: After 5000 miles I am actually still running on the SKF seal. It seems that the seal only leaks (at present) when the car is run at higher RPM. I suspect the heat makes the seal surface more supple and the coolant pressure is able to push past it.

Boxster EV wrote: Tue Jul 19, 2022 10:27 am I have now fitted the aforementioned SKF seal. It will run on the Speedi sleeve.

I took the car for a blast last night and everything seems fine.

I’ll update any relevant long term feedback in this thread.


Johan kindly did some calculations ref RPM and seal tolerances. With my wheel sizes, the seal will be in tolerance to aprox 12500 RPM which equates to 100MPH.

Porsche 986 powered by a Tesla large drive unit. Backwards. Build documented here and Instagram @tesla_porsche here.
User avatar
muehlpower
Posts: 570
Joined: Fri Oct 11, 2019 10:51 am
Location: Germany Fürstenfeldbruck
Has thanked: 12 times
Been thanked: 96 times

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by muehlpower »

I could also imagine a ceramic flat seal. https://www.kramp.com/shop-be/de/p/mech ... MS114030KR

Of course, this is too long in this design, but it should work with disc springs and/or a reworked housing. The speed would fit (20m/sec)
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

Boxster EV wrote: Sun Dec 04, 2022 12:01 pm Data point and update: after 3500 miles my SKF seal is leaking and progressively getting worse (I have a drain plug so it's dumping coolant overboard). I can see the expansion tank level dropping every few hundred miles and am having to keep it topped up.

I'm not going to have time to remove and examine it before Christmas, but I thought I'd post an update before someone follows my lead with the SKF route (IE don't). I'll take pictures once out of the car.

As a reminder, the SKF seal I used is actually designed for oil applications so it was a long shot expecting it to last any reasonable length of time.

Looks like I'll have to get my rotor shaft repaired and refinished (so that I wont need a speedi sleeve), however we're all still stuck on finding a decent replacement seal.
:( Looks like seal hunt continues... I'll be posting a lot more details to aid the community in search of a solution. Will follow up with details on FKM seal characteristics and shaft hardness info in individual posts. I think shaft hardness info matters in your shaft repair effort.
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

Shaft Hardness

When talking to seal developers, they all ask about shaft hardness. For wear longevity, its desirable to choose the hardest seal material which doesn't harm the shaft.

@WimV's rotor shaft was scored after using Tesla's triple lip seal ( viewtopic.php?p=46359#p46359 ) Perhaps THAT particular seal's material was too hard? Tesla's supply chain probably tried a lot of different seals judging from every PTFE seal maker seems to have heard from them.

Determining Shaft Hardness

Without the rotor shaft hardness test data. Don't really know. Rc specification (Parker seal catalog 5450 table 9-3 pg 8-8 https://www.parker.com/literature/Engin ... /5350.pdf )
stands for Rockwell C. https://www.yorksaw.com/rockwell-hardness/ Looks like involving spinning a material against known hardness material and evaluate the wear.

Rotor Shaft Hardness

This section was outlined by @Johan

It is likely the shaft material is not stainless steel (we see enough that has sat idle to be corroded and rusted) Rather, it is likely SAE/AISI 1045

https://sinotech.com/blog/choosing-material-for-shafts/
http://what-when-how.com/electric-motor ... ic-motors/
https://www.eng-tips.com/viewthread.cfm?qid=461308

This steel's production process can be either cold- or hot-rolled. It's assumed to be hot-rolled for less residual stresses

https://www.ctemag.com/news/articles/ho ... lled-steel

resulting in a higher precision as required for the seal. Hardening treatment options are Through-, Flame-, and Induction hardening

https://www.fushunspecialsteel.com/

For a 1045 steel, the maximum hardness after hardening is in the range of Rockwell C 54-60 (avg = 57)

https://doubleeaglealloys.com/products/ ... 045-steel/
https://www.azom.com/article.aspx?ArticleID=6130

So it could be assumed that the hardness is in the upper fifties of Rockwell C. To express hardness on a different scale, use the hardness conversion table

https://en.wikipedia.org/wiki/Hardness_comparison

Repairing Damaged Shafts

Looks like this is the process ( https://youtu.be/Cr78CdFyD9w?t=945 ) As this process adds material, would be interesting to know if Rc can be determined. Video doesn't show any step that might be nitriding ( I'm complete novice ) Turn on youtube close caption and auto translate to see what they are saying.

Guesstimating Shaft Hardness

Parker seal table link above shows list of minimum shaft hardness Rc. Without knowing the spec, its natural to bias towards assuming lower hardness shaft surface.
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

Face Seals

I used the word water pump seal loosely. Correctly described is face seal.

- One face is ceramic
- Other is carbon (wear surface).
- Energized by a spring pushing the 2 face together
- During operation, thin film of media lubricates between the face to avoid burn up. Inevitable minute leak to dry side need to be vacated via chamber, weeping hole and barrier to bearing corrosion.

Classical picture in water pump application here again

https://www.agcoauto.com/content/news/p2_articleid/190

Argo Seals ( https://www.ergoseal.com/ ) provided a CAD drawing (under NDA) proposal (2 years ago) to Tesla supplier designed to fit into LDU. No way to pay for dev+test to even answer key questions so no way to go forward

- Ceramic and Carbon material/grade spec can be chosen to fit application
- Found a tiny area for spring (not much space)
- Need to be assembled in 2 halves (one face side mount to manifold, the other to shaft)
- No engineering engagement to ask about drain channel routing. Ideally include internal leak routing path that mates to a drain hole we tap.

State of the art is the EKK video posted in prior posting ( https://openinverter.org/forum/viewtopi ... 858#p46858 ) With micromachine pumping aid structure and narrow width for EV motor application. Strictly OEM business volume design effort.
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

This seal design source deserve its own source. Seminal work on seal design eduction

Found online, downloaded and convert to PDF with key sections bookmarked on PTFE seals, FKM material, PTFE lipped seals, classical V lip seal design background

https://drive.google.com/file/d/1T_A4wh ... sp=sharing
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

PTFE Seal Shaft Surface Prep and Install Challenge

Shaft Surface Prep

Gleaming from various sources, proper shaft preparation seems critical

- noted by tech at QC Charger (I think in a PM)
- THE seal design book in above thread
- just plain old first principle

PTFE seal not going to work on pitted, scored, or probably even over thin layer of oxidized coating
PTFE seal not going to work with any kind of lubricant applied to shaft to get the seal to install (not the way anyways)

For surface prep, we have 3 sources

- @johan did 600->1500 grit prep over narrow band on coolant shaft where prior single lip seal polished. In his case, the Ceimin primary coolant lip would be engaging original shaft surface treatment out of factory.
- I did 600>1500 grit prep over entire 12-13mm shaft surface. Had to remove oxidized coating just beyond the original single lip seal on the wet side. 600grit took that off in a couple of rotations. Details here https://teslamotorsclub.com/tmc/posts/7123529/ (I put a lot of info on TMC thinking the advanced pros here don't need it. But the details on LDU are so diverse and tricky details specific to this seal is rather important so look over the thread at TMC for additional segments of the LDU rebuild)
- Norwegian LDU rebuilder use 2000 grit
- PTFE seal manufacturer in OK https://advanced-emc.com/selecting-the- ... aft-seals/

Note the Advanced-EMC page states

- Too rough = seal wear and leak path. But need enough to transfer a thin layer of PTFE material onto the mating surface
- <= 45Rc results in polished surface. Since LDU coolant shaft see some polishing. Perhaps its not quite 60+ Rc.

@Johan and I only have 500mi so far on this surface prep along with Ceimin triple lip yellow PTFE seal. So the experiment continues.

PTFE seal excluder lip install

This is really really error prone. excluder lip is 3+ mm narrower than 30mm shaft. It will fold if not stretched properly. Furthermore, stretched PTFE lip will narrow again within hours making it harder to install again.

The trick is to stretch it out and fit onto the shaft. Then immediately remove it and mount with sealant onto the manifold and seal the assembled manifold onto the shaft/endplate. Avoid waiting long delay in between for the seal lip to shrink. Following link includes full descriptions and attached PDF showing all the tips and proper mounting geometry.

https://teslamotorsclub.com/tmc/posts/7185566/

Changing Shaft Diameter

@WimV did a incredible job with combining 2 speedi sleeves to cover the entire shaft's seal surface and used Ceimin triple lip yellow PTFE. Unfortunately found beginning of a leak (at speed sensor) at 250mi mark

https://openinverter.org/forum/viewtopi ... 364#p46364

We can guess a couple of things

- Speedi-Sleeve changed diameter from 30mm to 31mm. PTFE is energized by the bent material wanting to bend back. Lathe cutting PTFE lip diameter surely accounts for shaft diameter and material (particular PTFE blend) stiffness. Don't know the consequences of stretching the material by another mm
- PTFE material is a blend and made via sintering process (same as ceramic balls). Basically ground into dust and heat until material combine together. Sintering process can leave interior voids and is source of fracturing. Lots of research papers on this.
- Chamferred Speedi-Sleeve edge at coolant shaft inlet bevel is impossible to achieve perfect mating. Don't know how much damage to the coolant facing lips on install. Coolant lips are even more narrow than excluder lip (4mm < 30mm shaft) Require a lot of force to push it in.
- 2x Speedi-Sleeve also reduce the maximum width of the shaft's seal contact area. Triple lip seal already have very small margin remaining on the shaft (~0.5mm each side) Further reduction could land seal lips on chamfer/bevel edge.
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

PTFE lined Seal

No one has tried this other than the snowmobile water pump seal that leaked at 40k miles on the website in my signature. But our information is imperfect, no idea if the leak was at speed sensor (leaks will show here first even before coming out of drain) or how much was coming out of the drain. Haven't been able to clarify this.

Seal Design

Parker's 5350 catalog ( https://www.parker.com/literature/Engin ... s/5350.pdf ) has a good profile on pg 4-6 in its Clipper-Sliptite Series. These include a layer of PTFE lining on the dry side of the V lip

The way I think this seal works is the asymmetric angle V lip is the classical common seal lip with its pumping aid (pump direction from shallow (dry side) angle to steeper angle (wet side)) The base material (say NBR) IS the energizer that puts pressure on the contact surface, The thin layer of PTFE lining just bends with the energizer polymer and act as a high surface speed high temperature lining.

Asked this to the Parker app engineer and they said V shape is just a reflection of the lathe cut. No pumping action. App eng didn't want to give me much time, trying to get off the phone and perhaps was referencing a different seal (100% PTFE cut to V lip shape which would be very difficult to bend as PTFE is so hard)

There are some factors which don't make sense given our lack of seal expertise. Elastomer (NBR, FKM) V lipped seal depends on shaft surface micro roughness to act like puddles on roads. Then hydroplane over them like tires. And with a tire that has high negative camber, sprays hydroplanes fluid towards one side. PTFE lipped seal works in much different principle so what exactly is happening in this case of a PTFE lined V shaped elastomer substrate seal is unclear.

I believe anytime we see a PTFE lined seal, it is likely this design profile. There are some available seals in 30x55 geometry. Here is one made by Taiwan manufacturer, https://sites.google.com/view/teslaldu/ ... mq6ldzyvzd

Longevity Hard to Ignore If True

If the snowmobile water pump seal longevity is correct and no significant leak level until 40k miles. It is hard to ignore this data. If anyone has better contact to Nathan and can get clarification (leak how much and where, speed sensor? drain port? collecting bucket?)
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

PTFE Lipped Seals

This likely is still the main target design to seek a quality source. Tesla and suppliers have repeatedly engaged most PTFE seal companies already and changed design design from triple to single lip which seems worse for leaks but maybe better for shaft wear? It could also be they chose too hard of PTFE blend material for the triple lip to operate without shaft damage?

Tesla is just nursing LDU cars out of its 8 year warranty period. WIth seals lasting couple of years before leak and noticeable damage another couple of years after, 2-3 LDUs shift this problematic design's repair cost to the customer's wallet.

With all of this background, I think we are just seeking a 2-3 year seal from a reputable source and can engage PTFE experts to make some good choices.

After much research and learning. I think there is not much magic secret. The key factors are

- PTFE blend. Need the longest lasting seal that doesn't damage the shaft. Since don't know shaft hardness, have to error on the side of seal being too soft

- Shaft need to have proper surface prep and can't change diameter

- Installation is tricky and require expertise

- Pumping aid. No need to search. None are available. State of art research is at Univ of Stuttgart for last 20 years. Research designs that mold microshapes and densities to create a pressure gradient. Reached out to lab professor, no reply. Looking at the amount of debris in Tesla's single lip seal with bending assist concentric rings, its not clear pumping aid will be all that successful. Regardless, industry isn't making any (even custom) with pumping aid. Classical helical grooves is unidirectional while we have a bidirectional (but mostly unidirectional) application.

- Lip geometry

Probably well defined formula based in shaft diameter AND material property.

====

So what is testing worth? Not that many changeable parameters. Given this, perhaps completely okay to skip formal engineering development and testing and just go straight to in car install. Realistically, thats what we are all doing anyways.

So now its just down to getting hold of a reputable source able to make small volumes and engage PTFE experts to make best material choices. We have the fixed geometry and shape.

Since Saint-Gobain and Parker Hannifin's custom seal group aren't really all that reachable. Time to go back to the smaller manufacturers who are usually specialized for medical and defense industries. Perhaps the first one I reached out to (Advanced EMC) is still the best option and get us a run of seals for ~$100 each. I'll explore and report back.

====

Ceimin could be perfectly good source given the above design criteria. Just difficult to reach their engineering to assess design skill and unknown brand. In my purchase experience, QC is questionable and not confidence building. Thus, a more reputable source could just be duplicated effort but many experienced machinists have learned to trust 1) reputable brand for QC 2) reputable distributor to avoid counterfeits.
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

Anyone has experience with isolators? In my conversation seal manufactures, I've heard isolators is what you want to really keep stuff out. Here is an animation of one



Isolators is basically a micro machine tiny labyrinth (leaks have to go through a tiny maze) with centrifugal force driving the leak to outer edge to an exit port. When its spinning, there is 0 frictional contact btw the piece that spins with the shaft (isolator rotor) and the piece that mounts to a bore (isolator stator). Any leak making through the labyrinth gets centrifuged out a controlled exit port.

Lets imagine the following design

- Use a single lipped PTFE seal. While these leak sooner than triple lip. The lip stays in pretty good shape (always have lubrication) and probably last awhile with minute leaking amount. Not even enough to get a coolant low warning or even noticeable in the reservoir.

- Can make single lipped seal cage narrower, freeing up space for an isolator just behind it. The rotor part of the isolator next to the seal. Stator part towards the reluctor wheel. Isolator exit port face towards ground... route an exit port on the manifold (avoid having to seal more mating surfaces for a weep channel on the endplate).

Result is minute leak makes it past the single lip seal that can probably last awhile then routed out by isolator to a weep port.
seal isolator.001.jpeg
I think it can even be simpler, somewhere between a full blown isolator and a centrifugal slinger is a solution. We are not talking handling much leak. Just need to sling it out in a controlled path and keep reluctor chamber dry. Nothing for the reluctor wheel to agitate, stick on walls all over and gradually attack bearing seal.

I'll look for PTFE seal manufacturers that also does isolator (seems reasonably common, after the same goal... prevent bearing destruction from leaks) and see if this make sense and can be made.
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
User avatar
Boxster EV
Posts: 434
Joined: Fri Jul 26, 2019 9:32 pm
Location: UK
Has thanked: 35 times
Been thanked: 26 times

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by Boxster EV »

Thanks for posting this information, Howard. The community must find a solution otherwise there will be a lot of LDU off the road in the near future.

I cant find a local link to the snowmobile PTFE lined water seal in the UK. Do you have further information on European suppliers? It only takes me an hour or so to change out the seal and would be willing to give it a try (on the speedi sleeve).

Additionally, do you have a view on whether Nitrile rubber will be more durable than fluoro rubber? The HMS5 S seal I am running is the latter and I'm wondering whether a nitrile equivalent is worth a try?!
Porsche 986 powered by a Tesla large drive unit. Backwards. Build documented here and Instagram @tesla_porsche here.
User avatar
Johan
Posts: 28
Joined: Thu Jun 02, 2022 11:59 pm
Has thanked: 2 times
Been thanked: 1 time

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by Johan »

Boxster EV wrote: Mon Dec 05, 2022 3:08 pm Additionally, do you have a view on whether Nitrile rubber will be more durable than fluoro rubber?
Nitrile is worse than fluororubber. See also the graph on page 6 of this thread.
Disclaimer: Despite all efforts, all (which I write) should be conservatively interpreted as a poorly informed, error-prone, non-expert opinion that is subject to continuous change, especially in this age of hyper-specialization and newly gained insights.
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

Boxster EV wrote: Mon Dec 05, 2022 3:08 pm Thanks for posting this information, Howard. The community must find a solution otherwise there will be a lot of LDU off the road in the near future.
Yes, indeed, On the seal hunt until Tesla send a cease and desist. Luckily I'm not youtubing against their PR haha. 2015 Model S comes off drive train warranty next year. Tidal wave is starting to hit owner's wallet.
Boxster EV wrote: Mon Dec 05, 2022 3:08 pm I cant find a local link to the snowmobile PTFE lined water seal in the UK. Do you have further information on European suppliers? It only takes me an hour or so to change out the seal and would be willing to give it a try (on the speedi sleeve).
I suppose my google results are naturally steering towards US so EU suppliers don't show up on my radar too often. Currently pursuing

- Chasing down the Parker triple lip PTFE
- Follow up with Parker to reach app engineering for the 3rd time and see if can gain better understand of PTFE lined design. Most off the shelf seals are from low end manufacturers (more info below) so engineering access is 0.
- Will reach out once again to smaller companies with good triple lip PTFE blends and speciality
- Will reach out to seal/isolator capable companies to see if +isolator idea works and is possible

As noted before, I'm beginning to realize dev+bench test while ideal may not be all that beneficial with limited degree of adjustable design parameters. And you Porsche EV guys might just be our savior with custom subframe with more easily accessible coolant manifold than Tesla's alien lifeform subframe. Spoke to another Porsche EV project yesterday. Stuck on coolant seal like everyone else. But quick change is possible like you :)

PTFE lined is interesting. Both in the single data point we have and theory behind the design. https://openinverter.org/forum/viewtopi ... 526#p49526 Perhaps @Johan can chime in as well

I have had a single DM with Nathan in September on TMC. He said started leak after 40k miles and he is now switching to yellow triple lip PTFE seals from Ukraine on ebay which we think is just Ceimin. Didn't hear back afterwards on multiple reachouts. Would have liked to get more precise info on how the leak was determined to understand if was caught early or late.

In terms of sourcing. Sounds like snowmobile water pump seal and other sources I found aren't any fancy manufacturers. All probably low cost seal manufacturers in Asia with a catalog bin offering to global distributors. $10 seals.

https://sites.google.com/view/teslaldu/ ... mq6ldzyvzd

If you want to try PTFE lined, maybe measure the diameter post speedi-sleeve and see if 30.5mm... Not sure if that will make a difference to 30mm to these type of seals.
Boxster EV wrote: Mon Dec 05, 2022 3:08 pm Additionally, do you have a view on whether Nitrile rubber will be more durable than fluoro rubber? The HMS5 S seal I am running is the latter and I'm wondering whether a nitrile equivalent is worth a try?!
NBR is worse than FKM as Johan said. I'll post a follow up on this outlining the science we learned on FKM and think is the reason for poor results.. If our understanding is correct, NBR will be same or worse.
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

Elastomer Seals (NBR FKM)

Lip Design Background

Like all seals, this seal relies on a film of lubricant between the seal and shaft to avoid burning up. To prevent this film from aggressively leaking to the dry side, All commonly available elastomer seals use a hydrodynamic pumping aid V shaped lip that has been around for many decades which creates a pressure gradient to pump the film towards the wet side

Image

Note air side angle of the V lip is shallower than lubricant side. When compressed onto the shaft, shallower angle will have more contact area with shaft than sharper angle side. This greater area over micro pot holes on the shaft forms the pressure gradient. pg 74-84 of THE seal text book goes over this with lots of pictures https://openinverter.org/forum/viewtopi ... 522#p49522

Contact Area and Friction Heat

This V shaped lip has a small cross section contact area (THE seal book quotes 0.25mm for example) This is where surface speed heat is the enemy. 0.25mm contact cross section x shaft diameter is the total contact surface area. All the frictional heat generated is on this limited surface with higher surface speeds resulting in higher heat density.

PTFE's flatter contact lip is like 2+mm between lip and shaft so 10x greater contact surface area to V shaped lip. Heat density is an order of magnitude lower besides a more heat tolerant material.

THE seal book goes over this on p84-85 (for V lip) and p96 for PTFE lip

NBR is the cheapest elastomer with the lowest thermal performance. FKM was developed for high thermal performance.

FKM (aka Fluororubber/FPM/Viton etc)

FKM was developed with higher thermal performance and handle a wider range of aggressive media. However it is weak to polar molecules (ie water)

SKF's thermal table on FKM shows same as NBR with water. p221 row (water) column (V=FKM) shows 100C (similar to NBR) https://www.skf.com/binaries/pub12/Imag ... 318140.pdf
The seal book also outlines this pg 33-35 https://openinverter.org/forum/viewtopi ... 522#p49522

I heard there are multiple grades of FKM. As noted previously, Parked developed a coolant resistant FKM. Its on the backlog of list of questions I have for Parker seal engineering whenever able to get some time. https://www.parker.com/literature/TechS ... G_5040.pdf
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

30mm vs 30.5mm with Speedi-Sleeve

The 30mm shaft Speedi-Sleeve spec shows 0.25mm thickness so shaft diameter will become 30.5mm after install.

https://www.skf.com/us/products/industr ... ctid-99114

Does this matter to 30mm seals?

How The Seal is Energized

Seal designer calls the squeezing pressure onto the shaft "energized" so need to understand the method it is energized.

For V lipped elastomers, its the garter spring. Parker will sell you a custom seal with shaft diameter in single mm increments. I'd imagine garter spring is just slightly different between each mm increment.

For PTFE lipped seals. The lip is likely cut to the predetermined diameter so material's stiffness property will energize with proper pressure over the shaft. Furthermore, Ceimin seal has no concentric cuts for "bending" aid thus probably less tolerant to 0.5mm diameter growth.

So... does 0.5mm matter? Just a guess but maybe not too much for garter spring energized seals? If this is true, then PTFE lined seals might be worth a try.

Love to hear what others think. @johan is more cautious than me on engineering analysis haha. I've been a manager too long!!!
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
smathermather
Posts: 12
Joined: Fri Dec 02, 2022 5:35 am
Has thanked: 7 times

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by smathermather »

I have confirmed that Mercedes-Benz is using a newer configuration of the Tesla LDUs in their replacements for 2017 and earlier B250e. This is the part:
https://web.archive.org/web/20221205200 ... 2340420184

And here is the part number: 242-340-42-01-84

I don't know much about it yet, but it is what will go into my bclass. Maybe MB has run down the need for a better seal and "solved" the problem. If so, finding out what they use for their seal might be useful, although I am unclear how one would find that info.
User avatar
crasbe
Posts: 234
Joined: Mon Jul 08, 2019 5:18 pm
Location: Germany
Has thanked: 43 times
Been thanked: 97 times

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by crasbe »

smathermather wrote: Mon Dec 05, 2022 7:40 pm I have confirmed that Mercedes-Benz is using a newer configuration of the Tesla LDUs in their replacements for 2017 and earlier B250e. This is the part:
https://mbparts.mbusa.com/oem-parts/mer ... 2340420184

And here is the part number for when that link inevitably dies some day: 242-340-42-01-84
The Web Archive has you covered: https://web.archive.org/web/20221205200 ... 2340420184
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

Parker PTFE lined Seal update

Called Parker EPS, spoke with sales support. Got a bunch of good info but looks like no go on Parker's PTFE lined seal listed on pg 4-6 of 5350 catalog https://www.parker.com/literature/Engin ... s/5350.pdf

Here is a summary

- Parker list all tooled up seals in appendix B (inches) and C (metric) EPS does mostly inches. Metric is mostly imported
- In inches (1.1811" = 30mm), appendix B shows no PTFE lined seals tooled up
- In metric (30mm), appendix C shows no PTFE lined seals tooled up
- No tooled up means custom. Now it gets expensive. I heard some rough #s and since Parker is large size company, its generally OEM business. > 1000 quantity, tooling up cost > $1k, $2k etc...
- To do custom, have to go through distributor. Support person named the following distributors supporting rotary seals (instead of exscvators! haha) Motion Industries, Applied Industries Tech, Kaman. Basically have to contact them, setup spec, get custom tooled up quotation.

This is likely the situation with Saint-Gobain's custom seal division as well if they were willing to accept the "proprietary" job haha.

Looks like big manufacturers are out. We are probably left with Taiwan seal manufacturers offering standard bin to small US distributors for PTFE lined seals. I'm sure thats what Kimpex was and the Suzuki and Honda part sources we found.

All of this is the same for triple lip PTFE seals from the big manufacturers. We likely can only source from smaller PTFE expertise companies serving industrial applications.

So while we'd like to have top brand names... They have no incentive to do small quantity special projects :(
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
howardc64
Posts: 166
Joined: Mon Sep 05, 2022 12:16 am
Location: Redmond WA
Has thanked: 8 times
Been thanked: 17 times
Contact:

Re: Tesla Large Drive Unit (LDU) Motor Teardown and maintenance

Post by howardc64 »

smathermather wrote: Mon Dec 05, 2022 7:40 pm I have confirmed that Mercedes-Benz is using a newer configuration of the Tesla LDUs in their replacements for 2017 and earlier B250e. This is the part:
https://web.archive.org/web/20221205200 ... 2340420184

And here is the part number: 242-340-42-01-84

I don't know much about it yet, but it is what will go into my bclass. Maybe MB has run down the need for a better seal and "solved" the problem. If so, finding out what they use for their seal might be useful, although I am unclear how one would find that info.
I'm guessing MB's LDU just comes from Tesla or Tesla rebuilders. I'd imagine Tesla has multiple rebuilder source now internationally. How many rebuild LDUs do we think they are installing daily? Here is the global sales figure of 12-19 at 260k

https://en.wikipedia.org/wiki/Tesla_Mod ... by_country

Lets say 130k are LDUs (50%) and LDU replacement average every 4 years (1400 days). Thats 90 LDU replaced daily. Maybe not all failed LDUs are rebuildable so lets say 3/4 of the 90 or 70 LDUs are rebuilds.

So probably 3-4 rebuild shops in US, 1-2 in EU, 1 in CN each rebuilding 10 daily. A tiny few will be MB and Rav4 EV with small differences.

Probably any new version of MB or Rav4 EV LDU are just Tesla's latest LDU revision #s. MB/Rav4 EV volume is much smaller than Model S so likely they will track LDU rev slower.
Tesla LDU rebuild website https://sites.google.com/view/teslaldu
2013 Model S85 3rd (Reman RevQ) LDU on 2017 @40k. Coolant seal leak discovered in 8/2022 @ 72k miles
Seattle US
Post Reply