Tesla Motor Coolant Seal: DIY Failure "Investigation" and Replacement 

You're welcome!

Yes that would be better but I did not have one...

Thanks for the input. We have done a ton of research about Teflon seals. See also the website in the description for more info.

Ik woon in California...

I did. See my other video on that.

@@diyEVguy Your right. You did cover on another video. Thank you.

That's a good idea! Thanks!

@@diyEVguy I really enjoy your videos on this topic, and I do not want in any way to be discouraging of your work and presentation. I, too, am not an expert, and I've learned that there's always somebody smarter than I. Below is mostly opinion based on my experience and education, and is offered not specifically for you personally, but for others who may be watching/reading. This sealing assembly is an interference fit, meaning that the seal's OD is slightly larger than the housing's ID, and the seal's OD is designed to deform and crush slightly upon insertion to the housing. Preheating a seal's housing (the manifold, in this case) can be an effective installation technique, especially when the housing is aluminum, due to the greater expansion rate compared to the seal's steel -- if both were the same material, heat would not reduce the interference, although it would make both materials more "fluid" in a relative way -- but the note you've applied at the top of this page mentions "250C", which is quite high IMO. There is both a safety issue -- this is quite hot to be handling -- and there's a possibility of seal damage from heat transfer. Sure, PTFE itself is quite heat tolerant, but the seal assembly may be composed of non-PTFE materials (for example: bonding agents) that cannot tolerate that extreme. Off the top of my head, I can imagine that the hollow rotor tube that looks to be steel, pressed into the aluminum manifold/cover/seal housing, may be installed using its own sealant (such as Loc-Tite) which is rendered useless at 250°C. Put another way, using heat at that level may introduce a problem while solving another. I would think milder heating, in the neighborhood of 150°C (~300°F) would be more appropriate. As I have access to a hydraulic press, I would be using a steel pipe adapter and press for installation, rather than the much higher shock load of impact installation. That manifold casting doesn't look robust. I liked your paper-tube-and-epoxy initial attempt, though. I've certainly beat enough seals into place over the years, but this isn't a wheel bearing seal here, where the stakes and tolerances are low. Lastly, as the volume of "air" -- a leakage path -- between the manlfold's counterbore and the seal's OD will be extremely low, it's not necessary to use even 5% of the RTV that you used in the counterbore; a mere "smear" of RTV would be adequate, and even then almost all of that will be pushed to the back of the counterbore and do nothing. The purpose of using a semi-liquid seal compound is to fill in any machining imperfections of the counterbore or surface irregularities in the seal's OD (or fill any grooves created during installation), so very little sealant is needed. Overuse of sealants is a common cause of surplus sealant eventually breaking free and ending up somewhere where chunks of sealant should not go -- such as smaller coolant passages or jammed in the water pumps' impeller and/or inlet ports. Again, Thanks for this and your other excellent videos. I especially like your phrase, "Do better", as that sets the proper expectations :)

@@alsavage1 Thank you so much! I will update the descriptions accordingly. You have very good points.

@@alsavage1 Thank you for the added insights. I modified my description accordingly.

@@alsavage1 What do you think about using anaerobic sealant in place of the RTV?

62k in total or only on the new DU?

@@diyEVguy 62k in total on the car. I talked to someone yesterday at a supercharger who has had his DU replaced 4 times now with 115k miles =(

@@eastmanresearch3143 Oh, you are just getting started. This guy replaced his DU 13 times in a million mile: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-53E2vAF5DtA.html

@@diyEVguy I wonder what the permanent fix would be. It's ridiculous that Tesla can't make these last 100k. I wonder what the root cause is in most of these situations - main bearings or coolant seals - I would think the fluid change interval of the coolant and/or differential fluid would get some contaminants out that might wear out the seals. I'd be up for doing coolant and differential fluid changes every 10k miles if it doubled the life expectancy of the drive unit. Whoever can come up with a fix for this outside of Tesla will make a fortune.

@@eastmanresearch3143 Judging by the posts of QC Charge, the coolant leak may be a large fraction of the DU-related problems. But common mode voltages, combined with agressive driving, causing fluting, may also be a common root cause. Personally I would try to avoid frequent large accellerations.

250mm in my case total length

​@@diyEVguyGreat, Thank you!

Correct, thank you, I should make a new video about that.

Yes it's a rotor coolant delete manifold. QC Charge used to offer them and rumor says that Tesla is now using no rotor cooling on RMN LDUs. However rotor temperature and the grease melting out of the bearings might be an issue.

See link below video - I just added it.

Oof! Good luck.

Good points, parts of those are discussed in the link in the description.

if they gave you a classic Tesla RMN motor, it will happen again. They last max 150-200.000km.

@@peterhandke3936 It did after 120.000km. Caught it early. Found red shaft seal. Installed drains and 3 lipped chinese seal.

Thanks!

Dutch man in the states? Groeten!

@@jonitan3584 ja. Not sure if milling is related to a leak. Could be that you have a leak and bearings have been flushed dry and are worn out so that the gears start to grind, but that's all quick speculation.

@@diyEVguy Dank je! ik denk inderdaad dat dat gebeurt is.

Oils have higher viscosity than coolant so it will run much hotter probably which is not good.

The coolant now used in Teslas (and other EVs) is G48 which is non conductive.

Because of the reasons outlined in this lengthy thread: openinverter.org/forum/viewtopic.php?t=2242 So if you know a better commercial off-the-shelf alternative, the community is open for recommendations 🙂

See also here for more replacement parts info: sites.google.com/view/teslaldu/parts

Corteco is also Giant in sealing industry provide OEM for lot of car manufacture@@diyEVguy