2000 Honda Insight - Heating the manual trans with engine coolant

[Daox]

Quite a while ago I started an epic thread on heating a manual transmission that included much research on the benefits of it. I mean, automatics do it, so why shouldn't manual transmissions? Traditionally they're air cooled and really probably don't even heat up that much on short trips. But, what are the benefits of having a warmed up transmission? How much energy is lost to churning thick fluid? I live where it gets really cold in winter, this only magnifies the problem. The worst case scenario that I could find said that there was a 15% improvement in fuel economy on a short trip with a pre-warmed transmission, and I think the temps were on the low side too if I remember. I really am not expecting that kind of an improvement. My normal commute is 7 miles, so my normal trips are quite short. I think its worth a shot, and it'll be fun to see a new mod around here.

Since I am pulling the transmission on my 2000 Honda Insight, I thought it would be a perfect time to give this modification a try. The idea is pretty darn simple. I tap into the heater core lines, tee off a new circuit to flow coolant to the transmission to warm it up. The transmission will have a jacket added around it, and the warm coolant will flow through this jacket. Thus, the warm coolant will heat the transmission case and also the fluid inside it. This method IMO is the easiest one I've come up with, and it also ensures that there will be no trans fluid to coolant being mixed.

To get things started, I ordered myself one of these earlier this week. It is a heater control valve and is typically used to shut off coolant from going to the heater core on a car. I'm going to use it to control the coolant flow to the transmission. I do not want to slow the warm up of the engine because of this new loop, so the new loop will only be activated once the coolant has warmed up to around 140-150F.





In order to actuate the valve, I plan on mounting a servo to it and letting an arduino control the servo so it is all done automatically. I already have an arduino in the car that is reading my coolant temp and controlling my automatically actuated grill block, so adding this extra servo and controlling it will be no big deal.





Last night, I got the valve in the mail. It was an inexpensive Four Seasons 74828 heater valve from amazon. It comes with some provisions for mounting a pushrod, but the servo will mount directly to it as shown above. So, I 3d printed a bracket for the servo last night. I still have to make the pushrod and nab a servo, but then this part will be done.







The next step will be modify the transmission by adding the coolant jacket around it. This will be an interesting step and I'm not 100% sure how I'm going to do it quite yet. I have to get the trans out of the car first too!

[rmay635703]

Could you stagger your coolant loops so the radiator loop opens at a higher temp?

That way you could test if the xmsn cools the car enough by itself

[Daox]

The transmission will cool things a little bit, but I don't think it has nearly enough surface area to really cool the engine down... Of course, it is pretty easy to cool the engine down in the Insight in winter by blasting the heat. In any case, the trans warmer will activate at a lower temp than the thermostat, so we should be able to see how it effects the coolant temp. That set point is fully adjustable, so I can play with it. I can also open the heater valve variably to get a little flow or a lot of flow. Lots of variables to play with.

[oil pan 4]

I wouldn't try to put a coolant jacket around the transmission at all. Transmission cases are die castaluminum. Die cast aluminum is nearly impossible to weld with out cracking it.
If I were going to do it I would replace the fill and drain plugs with 90° barb nipples and pump the oil to a heat exchanger and through a filter too then back to the transmission with a marine oil pump.

I will likely be doing this on my firebird six-speed transmission at some point.

[Daox]

I completely agree that welding to the aluminum transmission case is a bad idea. That is why I will be using glue. I'm not sure what type of glue yet, but cars these days are mostly glued together. I'm sure I can find something that will hold up to the ~15 psi that most cooling systems are designed for.

[ThermionicScott]

I like the idea of using the drain and fill plugs as connections!

Maybe less exciting of an idea, but what about an unpumped loop between the transmission and a coil of tubing in the engine oil pan? It would allow some exchange of heat as the transmission fluid sloshes around during driving, but not rob a lot while the engine is still warming up. And would be really simple.

[ar5boosted]

Originally Posted by Daox

I completely agree that welding to the aluminum transmission case is a bad idea. That is why I will be using glue. I'm not sure what type of glue yet, but cars these days are mostly glued together. I'm sure I can find something that will hold up to the ~15 psi that most cooling systems are designed for.

fwiw, you can use silver-solder to join Aluminium successfully. I actually just gave a course in Japan on using Durafix which is a type of Aluminium Solder for want of a better description.


[My Aluminium Rearmount-turbo-intercooler pipe]

You can find durafix on ebay or somewhere. I made some Aluminium computer-cases https://forum.makehackvoid.com/t/diy...-projects/1171 just to show that it can be done.

It requires inexpensive tools like a propane burner to heat the base metal.

See this link to youtube for other examples : https://www.youtube.com/results?search_query=durafix

[Daox]

Wow, that durafix looks like nice stuff. I love that you can build it up to fill gaps. That'll make fabricating the outer skin a lot easier and quicker. Thanks!

https://www.youtube.com/watch?v=kMvuwS3e24A

[pete c]

I assume a CVT insight has a tranny cooler built into its radiator. Why not get one of them and pump the tranny oil through it?

[Daox]

While pumping oil through a heat exchanger is a great idea and may very well work better overall (because it is directly heating the trans fluid). I would like to stay away from it simply because it is additional wear parts and more complexity. I prefer the method of just using the existing cooling system and adding a valve because its relatively simple. The simpler, the more reliable, the cheaper and easier / quicker it is to make.