NHW11 Traction battery autopsy

I too was concerned about transaxle temperatures so I watched them like a hawk the first time. But nothing happened.

MG1 is rated at ~18 kW but when running as a generator, ~92-95% of the worst case, ~16.2 kW will be going to the battery. So we're looking at about 1.8 kW as heat. In normal Prius operation, power generated by MG1 goes to MG2 (and in something called "energy recirculate mode" from MG2 to MG1.) Notice in the graph that the initial peak charge is ~320 V x 45 A ~= 16.4 kW. Not quite perfect but close enough for engineering work. <grins>

To us, this seems like a lot but in reality, these are well within the heat cooling capability of the transaxle . . . but it is exciting the first two or three times.

I believe a forced charge is putting more stress on the traction battery from the exothermic, charge heating and potential for generating H{2} and O{2} gas. Internal battery module pressure is the enemy since that is what stresses the seals and leads to leakage. However, it has yet to be associated with traction battery failure. . . . I only do it for a specific reason such as diagnosis or measurement. Once I have my data, like a prostate test, no need to repeat that test!

Bob Wilson

 

Should MG2 *always* be less than MG1 or is this just a general case?

Out of interest, I checked the temps on my car and under normal cruise, MG1 > MG2, but when I hit some traffic and was doing stop-and-go, with a fair bit of low speed driving on battery, MG2 became > MG1; only by a couple of degrees, but it was hotter.. Once we got going again, the temperature reversed and MG1 > MG2..

 

The general case. What happens is ~28% of all power from the ICE has to pass between MG1 and MG2. But the traction battery has about 20 kW of power that can be fed from MG2 or worse, feeds MG2. If your car is in a mode where there is significant power flowing into and from the traction battery, MG2 can get pretty hot . . . and the vehicle mileage will be not so good. FYI, it isn't very good for the traction battery either!

When my wife was driving to Nashville, she maxed the car heading up the hills which sucked power from the traction battery. Then on the down grades, she kept the speed up and the ICE was generating power to recharge the traction battery. It was horrible but she can't do that with her ZVW30. The 1.8L engine solved that problem!

I can model extreme stop-and-go driving that would accomplish the same effect. "Top Gear" did that with a 17 MPG Prius run.

Bob Wilson

 

No, that's not true. MG2 can run hotter than MG1 during normal operation. Let's not scare people.

 

It has been a while since I last did a forced charge and I've got Auto Enginuity which records more data points than my Graham miniscanner:

The driving profile:

• 0-300 seconds - normal cutting through neighborhood to warm up engine with minimum fuel burn. Stopping at light before getting on cross-town, Parkway.
• 400-1050 seconds - mostly using the access road, posted 40 mph with GPS speed showing 40 mph. However, the indicated is closer to 38-39 and the Auto Enginuity recorded speed even lower.
• 1050-2000 seconds - cruising at GPS 65 mph, indicated 62 mph, and OBD reported 60 mph.
• 2000-2100 seconds - turning around, regenerative braking puts a charge on traction battery and MG2 climbs as speed was in EV range. The engine and MG1 took a break.
• 2100-2700 seconds - running at GPS 75 mph, indicated 72 mph, and OBD reported 70 mph.
• 2700-3000 seconds - driving to work parking lot
• 3000-3300 seconds - forced charge. In the past, I'd only looked at MG1 temperature but this time, I also monitored MG2. I was surprised to see once the traction battery reached 80% SOC, MG2 took a distinct rise while MG1 stayed constant.
• 3300-3900 seconds - left car parked, in READY and went into work.

I've got a healthy transaxle and until I parked, thus stopping the forced air through engine compartment, MG1 was warmer than MG2. So it looks as if someone is at exceptionally low speeds where there is little cooling air flow, MG2 will approach MG1 temperature. But it really takes significant power, regenerative and EV, to get MG2 warmer than MG1. A forced charge will definitely warm-up MG2, something I hadn't seen documented before.

BTW, I did a variation of the EV hill climb using feet on floor descending of a test hill. The SOC went from 66.4% to 68.75% with starting and ending speed of 25 mph. The mass of the car and change in altitude gives the energy needed to add 2.35% SOC. This should give a reasonable estimate of the traction battery capacity. For grins, I'll try a taller hill to improve the accuracy.

Bob Wilson

 

Sounds like a good plan. You may want to get the old buss bars ready and/or get a new set. Regardless, it looks like an excellent plan and you've pretty well nailed all we can know short of disassembly.

Bob Wilson

 

I actually set to 2 decimal places but somehow SGII truncate it to 1 decimal place. Anyway, if you remove the decimal place formatting in RXF, you can get the 2nd decimal places:

	Column 1	Column 2
0	RXF original	03D5444105A4
1	RXF modified	03D5044105A4

IMHO, there is a safer alternative to do this. The repair manual describes a stall test:

 

I have in the past done the decimal point RXF edit for SOC xgauge, and it worked thanks.

 

When you get a chance, could you check your old modules to see if there is any evidence of surface discoloration? You may need to look at them from an oblique angle:

. . .

I especially appreciate the thermal graphs of the old and new style traction battery modules. I'd seen postings about the difference in the past but this really brings it home. It also explains the melted "O" ring material found at the terminals and the trends seen in NHW11 battery failures:

• warmer climates - higher failure rates because the traction battery is already starting hot
• hills - the best way to get peak charge and discharge rates

It is great to see hard data, empirical measurements, being brought to the forum. This also explains a curious pattern I'd seen in some of the old modules:


I had noticed the modules had a curious triangular, surface pattern which parallels the thermal model showing hot-spots. Literally the modules I picked up have been 'cooked' and the smooth sections are where the heat has released the original stresses on the case.

Thanks,
Bob Wilson

 

Just bought 3 lots of these used batteries (33 total) off ebay in hopes to make a semi reliable battery pack for my ebike project. Ordered extra to weed out the weak/bad cells. I just ordered a 989 charger last night.. I've needed a good battery analyzer for a while anyway. The batteries are showing obvious signs of leakage or sealant (tan colored material that is stuck on good).. hoping I can take them outside and pressure wash them clean. I will start with a 3 cycle 1A charge/discharge test to get a baseline and start a spreadsheet like you guys have done. I'm curious about the re-hydration.. do you only need to drill a small hole in one end and tilt the battery so the water will cover the tops of all the cells? Will it get to a point where the cells are saturated and wont take any more water? JBWeld seems like it might be a bit too runny to reseal the hole.. anyone tried scuffing the sides around the hole and using the plastic weld putty? Wish me luck

 

I wanted to know if anyone here has experience using the HiTech 4X AC Plus Charger. A link to the product is below:

X4 AC Plus Multi-Charger - Hitec RCD

Instead of buying multiple chargers I thought it would be better to purchase one that could charge multiple packs at one time.

Any input would be greatly appreciated.

Thanks,

Mylar

 

I got one like it about two weeks ago when the stealership wanted $4000.00 to fix mine. Hmmm

It worked great! I was able to charge 4 modules at a time. Get a good 12v power supply. I used a battery on trickle charge.

Just got a scanguage 2 as well and learning how to use it from this site. They also wanted to replace the MAF sencer $300.00. A $5.00 can of cleaner fixed it right up. Hmmmm

Thanks to all for the help from this site. Batteries back in with one replacement module and things look OK. Battery seems good for now and I am keeping an eye things.

 

Did you use the auto function or set the charge / discharge parameters manually?

Thanks,

Mylar

 

Since my knowledge is limited I let it chose the charge limit. Up to 8.3v on all and then it would stop and discharge. I allowed a 2.5amp charge, 1 amp discharge, and kept them in the rack to minimize expansion. I only allowed it to discharge to 7v. Not sure if I should have went a bit lower. I ended up doing them individually, 3cycles and then in parallel dividing the bunch into 4 parallel groups still in the base 3 cycles. I could show you the wiring I did to do this. Quite simple really. From a simple guy come a simple idea! I monitored the voltages closely and only joined it all together again when the voltages were the same and removing the one dead cell. Would or should I have done some things different possibly but so far so good. One guy on here just charged a cell and replaced the bad one and it seems to have worked fine for him so far.

Results from the Scanguage today are:
Bank 9 is my lowest consistent voltage followed by bank 5 sometimes.
A .2 to .4 volts different from lowest to highest.
Voltage varies but when its not charging the bank voltages all sit around 15.1v to 15.3v
SOC is in the 270 to 300 range.
Around a 60% 65% charge on battery (could be better)
Battery temp good
Can't wait to see how it fairs in the cold weather we are having.

Am I affrad to work on the car? Not a chance!
Do I know what I am doing? That remains to be seen!:rockon:

 

Photo of set up.

 

Just dropping in for a moment to post the outcome of the work I did on the 2001 Prius. I still have it and it is running quite well. No traction battery problems to date. I carry a quick fix package with me just in case I have a problem on the road. The SOC has been around 68% the consistant lowest batteries have been pack 9 at or around 15 or so volts. The car likes to give a misfire code every once in a while. Toyota must be making a killing with such sensitive cars knowing most people are afraid to touch them. That's all for now.

 

How did you arrange all these on the table? Want excellent.