Kit car (P)HEV conversion

[oliverracing]

Apologies if this is in the wrong area, if so I can move it. Long time lurker, but now starting to look at starting to build up a spec for this to hopefully build in the couple of years or so...

I have a RWD rear engined kit car, loosely based around two front rover metro subframes bolted to the front and back of a fibreglass tub (GTM K3 for those that are interested). Having a front subframe that originally had an engine in, there is plenty of space to add a small electric motor, plus the hubs are even designed to take driveshafts! The car currently only weighs about 850-900kg so want to keep as light as possible.

Something I've been thinking about doing for a long time is adding a smallish (40-70kW) as light as possible motor to the front wheels to allow more spirited corner exits, traffic light getaways or the ability to drive quietly as the 1.8l petrol engine is a touch loud for 2am village driving. I'd keep the battery small but hopefully big enough to give a few miles range if needed. Initially the idea would be to have it as a standalone/regen charging only setup but should the pack have enough capacity to be worth it (15 miles range I'd say) I'd look at adding the ability to charge from the mains.

In terms of motors I've looked at the rear drive unit from the Outlander as it's a nicely contained package that should bolt in fairly easily, but I'm struggling to find the maths on what output I would be able to expect if running at lower voltages - any pointers here?

On the subject of lower voltages, my plan would be to have a small pack for keeping the weight down but something that will allow me ability to pull high current from it for a short period - I'm not overly bothered about high speed as the existing motor will provide the power there so power dropping off above 35-45mph would be absolutely fine as long as I'm not overlooking something if I end up running the motor faster than the battery could run it? In theory the car can do 140mph so I don't want to end up with a battery on fire if I end up doing that on a track!

Also I know the lower voltages will limit my inverter choices severely, what options are there that will allow me to run lower? I can't find anything other than the volt inverter (might be wrong there too?) that will do sub 200V but I might be missing something?

Thanks in advance and sorry again if any of these are easy answers - I'm obviously googling the wrong things if so!

[johu]

Power drops about linear with voltage.
PHEV packs are low energy, high power. Would be a good base. See here https://openinverter.org/wiki/Batteries

[Pete9008]

Welcome

Interesting project, really like the GTM's. Is yours the 1.8 VVC K-series? If so it must be pretty quick even without the planned EV boost!

I'm currently doing a conversion using the Outlander rear and have been doing a fair bit of work in trying to understand the voltage/power trade-offs so may be able to help.

I also have a rwd track car and have been thinking along similar lines for quite a few years. There are a few issues though which is why it's something I'm still only thinking about.

The first issue is weight, the K3 is a pretty light car and adding a significant amount of weight it going to upset the handling. The outlander rear is one of the lightest water cooled transaxles and it comes in around 60kg, the inverter will add at least another 10kg, the battery anything from 20kg to 100kg so you are likely to be adding at least 100kg to the weight, possibly a fair bit more.

Next is batteries and this is the big one for me. Lithium batteries have limits on both how quickly they can be charged and discharged. I'm still looking for good reliable information on this but have settled on 10C as a limit for my use. My plan is to use a BMW 530 pack which are 34AHr capacity (so 10C is 340A). I'm using a full pack or 6 modules (~100kg) so around 350V so combined this will allow a max discharge rate of 125kW. Now if you halve the pack voltage to the 200V mentioned this drops to 70kW but with only 50kg of battery. It's also worth noting that track use will be very hard on batteries, the continual high charge and discharge currents can cause cell heating so cell cooling would probably be required. With Lithium you need to have plenty of capacity to get the high currents. A couple of people do claim to have a solution for this but to date I haven't seen any details. Another option is the NiMh packs used in some Hondas which are supposed to give good power density (higher current) but lower energy density (less kWh); these may be more suitable here but I haven't worked out any details.

The last is field weakening. Essentially motors like the Outlander's are permanent magnet based. This means that they act as generators too. If the motor speed rises above base speed they start to push power back into the battery unless the inverter prevents it. The base speed varies with battery voltage and for the Outlander motor is it something like 4.5krpm with a 360V supply (45mph ish). If you drop the voltage to 200V this frequency drops to 2.5krpm (25mph ish). At 140mph the voltage potentially generated by the motor is quite scary! Now the inverter should be able to prevent this but unfortunately the OpenInverter code is still a little iffy here, it's a lot better than it was, but running at 140mph would probably be a bit too much to ask at the moment.

An induction motor based transaxle like the Tesla SDU would make a lot of sense here as it is better able to handle the wider speed range. Unfortunately it's expensive, larger and heavier but might be worth a look?

Sorry to be a bit negative on this but I think the above is realistic. As mentioned this is a problem that I would like to solve too and I'm keen to explore possible solutions.

[Pete9008]

Been vaguely wondering about another idea for a while but prompted by your post finally got round to working out some numbers.

Say you use a pair of MG2 motors out of a Gen2 Prius transmission then according to the plot here viewtopic.php?p=48301#p48301 each motor should give 400Nm so a total of 800Nm. Say the specs can be pushed a bit more to get this up to 1000Nm for the pair (highly likely, probably even a bit more). If these were directly coupled to the driveshafts then motor speed would be limited to around 2krpm for 140mph which should avoid problems with field weakening. 1000Nm with 600mm diameter tyres would give f = 1000/.3 = 3333N at the contact patch and acceleration with a 800kg car of a = F/m = 4.2m/s or around 0.4G. Not great but not bad as a power boost and probably not far off the limit of what the front wheels could do on a car with a rear weight bias. It's also instantly available for accel or regen from 0mph all the way up to top speed.

I did wonder if the power split device could be kept to increase torque a bit more but not sure it can.

If you wanted to get really clever the toque for the motors could be individually controlled to provide torque vectoring to increase stability while braking, accelerating and around corners!

Big machining job but to date it's the best idea that I've come up with for a lightweight and compact PM based solution that could cope with the high speeds encountered during track use.

Edit - just checked a couple of things. First it looks like the power split device could be used to increase torque by around 25%. Unfortunately I've just checked the Oak Ridge report and a single MG2 alone is 30cm diameter by 20cm long and weighs 45kg! That's too big and too heavy for me.

[oliverracing]

Wow, thanks that's a lot of really great information and confirms a lot of my suspicions about the potential limitations of this project, but also gives me some ideas on routes to go down. Will fully re-read and digest later as that is all really useful and thanks again for compiling all that together!

Yes, it's a 1.8vvc - slightly tuned so about 170bhp and yes, it's pretty fast. I'll be entirely honest, a large driver for this project is for the sake of learning/why not as the car is a constant project and I enjoy tweaking/upgrading/modernising it. I'm also an engineer by trade and learning through doing is my favourite way! One route I might potentially go down longer term is fully electric, but that would be putting a large motor in the rear and bigger battery etc and something for 10+ years time.

In terms of battery, I've got a fairly decent space under where the fuel tank sits behind the seats, and potentially a bit of space in front of the center console, between these I reckon I'd have space for about 50-80kg of batteries here with a bit of work could probably fit a few more and that would be able to keep the weight distribution similar to current, so something along the lines of what you say would work and could give me a decent amount of capacity to play with. Currently running MGF running gear mostly, and they're 1070kg so want to keep under that, but that does give me a fair chunk to play with. I definitely want to maximise short (sub 10s) output even if that means adding a few extra kg and cooling wise I'm happy to severely limit the long term output as it isn't a track car and probably never will be.

The Prius MG2 idea is not a bad shout, not sure where the 45kg Comes from, maybe including the full casing etc? There's a bare one on eBay - https://www.ebay.co.uk/itm/184892115756 and the seller has confirmed this weighs 13.4kg as it is here. Yes it needs a housing but I reckon that could be a very viable option. Space wise it's just luggage space I'll be losing and happy to sacrifice that here and could easily package a couple in my application - with a bit of thinking I reckon 20kg for a case that holds two motors plus mounts is feasible, but yes would be a big machining/welding job but not beyond the realms of doable.

Will do some more research and see what I come up with!

[Pete9008]

No problem, it's something I've been thinking about for a while.

Actually started building a Libra 20yrs ago (same VVC engine), unfortunately moved house before finishing it, life got in the way, requirements changed and ended up selling it part finished. Still regret it in many ways but you can only have so many projects!

My track car is a FWD car that has been converted to RWD with a space frame chassis underneath. It still has the front drive hubs and a space of around 30cm x 30cm x 40cm between the front wheels where the fuel tank currently is. I'd like to come up with a solution that would fit in there, as you say to add regen into corners and a boost out. Longer term I'd like to put a turbine in the exhaust running a generator to allow some help on the straights too (maybe 40-50bhp?). I do want to keep battery weight down though, really just enough to hold 150mph worth of energy from a 1ton car

The weight came from the Oak Ridge Report on the Prius, see https://www.engr.uvic.ca/~mech459/Pub_R ... 890029.pdf page 49 and 50. It does sound very heavy though, your number is much closer to what I was expecting and it would be great if it is that low. Hoping the Ev8 will post the size/weight of his rx400h MG2 motor as another reference point (see viewtopic.php?p=53432#p53432). Need to be a little careful with parts though, the later Gen3 MG2 is lighter but designed to run at higher speed with lower torque and so would be less suitable. Looking at the design of the motor a housing shouldn't be too bad to do in terms of design, just a lot of fabrication/machining.

Edit - see viewtopic.php?p=53480#p53480 for Ev8's measurements and weights on the rx400h motor. It looks a possibility, little heavier but will probably generate more torque than the Prius one.

Edit2 - The main question is probably your desired top speed. If it's not going on track and you were happy with nearer 100mph then the Outlander rear would be a much easier solution (I think the spec for it is 106mph max, I'm not planning on using it much above 70mph and simulations so far show problems above 8krpm)

[Pete9008]

Got curious about what might be possible so decided to do a few simulations based on Ev8's rx400h motor data (the best data currently available). First plots still have a gearbox ratio set in the simulator as it allows the simulation to run a bit quicker, it doesn't affect the power or torque figures quoted as these are at the motor shaft. Edit - If I have my sums right then the 2krpm max on the plots, if used for direct drive, should be around 140mph.

First MG2 at 360V:

Pretty good, 400Nm to 1300rpm (around 90mph) and 60kW peak per motor so a total for the pair of around 160bhp!

Still MG2 but dropping the voltage to 200V:

Much faster torque drop off, still giving 400Nm to 50mph though so could still be worthwhile.

Also decided to see what MG1 can do as it is much lighter. First at 360V:

Now this surprised me, it needs a little more current to do it but it's making the same torque (and keeping it up to a higher speed) and even more power, a total of almost 200hp for the pair!

MG1 but dropping the voltage to 200V:

so again pretty similar to MG2.

Finally decided to simulate what a pair of MG1s at 360V could do in a 800kg car. No point showing the plot as it looks the same but it shows the car accelerating from 0 to 150mph in 20seconds. Torque is completely flat to 100mph and is still at 650Nm (for the pair) at 150mph! It would need a battery able to supply 800A though!

I suppose the big question is whether MG1 would melt or whether the cores would saturate at this current level. This might actually work though - I may start looking for a couple of rx400h gearboxes for when I'm finished with the current conversion.

Edit - If it would take 500A then simulation says that one MG1 could produce 650Nm The core is likely to saturate before this point but wonder at what level?

Edit2 - thinking about it these torque levels for MG1 would be nice but I think it is highly likely that the core will saturate or windings melt before getting there, MG2 is a much better bet even if heavier.

Edit3 - Before anyone else notices I forgot to set Icrit to the correct value in all the above plots Fortunately it's not pushing far enough into field weakening to matter so doesn't make much difference!

Edit4 - Fairly sure that the 13.4kg ebay weight above is for MG1 (the generator not the motor). Both the Gen2 and Gen3 Oak Ridge reports are pretty consistent about the weight of MG2 - 45kg (with around 10kg of that being the casing). In particular Table 2.7 of the Gen3 report contains a lot of useful data on the Prius, Camry and LS600h motors.

[Ev8]

Wow shame my gen2 Prius inverter won’t let me push more than 240amps before mg1 inverter trips!

[Pete9008]

Agreed! Although I think it will saturate before it generates decent power. Any idea what the Lexus spec for the 400h MG1 power is?

Just googling planetary gear sets to see if one could be grafted onto MG1 to let it spin a bit quicker. A Lexus RWD transmission would be ideal but they are a bit pricey. Wondering if a gearset out of a cheap autobox could be an option?

[Pete9008]

Been thinking about this a little more and I don't think the dual motor idea is going to work, not for track use anyway. First MG1 at 400A, and with a motor resistance of 100mR, means 16kW dissipation! Now efficiency isnt that important in this application but it's going to overheat massively. Possibly ok for a short burst but not for sustained use. MG2 at 300A would be better at 9kW but still difficult to keep cool and also heavier. A planetary gear set could be added, and the voltage increased, to allow the motor to spin faster and run more efficiently but it adds weight and complexity.

Still like the idea but more work needed.

For a road biased car a low cost approach would be to accept the lower top speed and use something like the Outlander rear motor. It will actually give a bigger acceleration boost at low speeds too. As mentioned above an ACIM based transaxle would be worth considering too if higher speed operation is essential.

[celeron55]

If it's possible to just fit an outlander rear drive unit, inverter and a BMW PHEV pack, I think that would increase the chances of success to just about 100%.

[Pete9008]

If it's possible to just fit an outlander rear drive unit, inverter and a BMW PHEV pack, I think that would increase the chances of success to just about 100%.

Yea, but where's the fun in that

Seriously though, if I'm fully confident that something will work I tend to lose interest in doing it, the challenge disappears!

Can't seem to give up on this so have given a little more thought and now believe it can be made to work. I'd dismissed the epicyclic gears because on the Prius MG2 drives them on the ring gear which is wrong for reducing speed (although I now believe this is only the case for the Prius, the rx400h is different having a second set of gears). Plus MG2 is already too heavy and adding the gears makes this worse.

Now I'd previously dismissed MG1 because I didn't believe it could do the job but I've now changed my mind. Moving to MG1 actually makes it all easier! MG1 already drives the sun gear and so the gears work perfectly for this. Using Prius ratios (I don't know the rx400h values but probably similar) it would be possible to get either a 2.6 or 3.6 reduction ratio depending on whether the ring or planets are locked. The 2.6 would be best here. The lighter weight of MG1 also means that the gears can be added without it all getting too heavy.

So chucking this into the simulator with a 360V battery gives:

Now the torque is before the gears so would increase by 2.6 giving 500Nm (per motor). The scaling is such that the 5.2krpm is still around 140mph so this full torque will be available to around 70mph, it will then drop to half this by 140mph. That seems a much better compromise. Motor current also drops to 250A giving a motor power dissipation of 6kW (thinking a little more, because of the way the calculator works out Rs, some of this will be in the inverter) which is easily manageable continuously. Power is around 75bhp per motor so 150bhp boost in total. This looks just about perfect for a FWD assist. A pair of rx400h MG2's driving through the gears would probably make a fairly potent stand alone RWD power unit with built in LSD and torque vectoring technology too!

The FWD assist looks like it might be my next project once the Smart conversion is finished

Edit - just for completeness simulated a run on a 800kg car, 60mph in around 6.5sec and 140mph in 18sec (ignoring drag so completely unrealistic though)!

[oliverracing]

All very interesting and will potentially be a route to get a lighter setup with some thought and effort, but also likely a lot more money.

I've been doing the maths (at a far simpler level currently) and tend to agree that whilst the MG1/MG2 ideas seem nice, as it stands currently as soon as you take into account the need to make a case, gearset etc you're going to end up over 25kg a motor - and the outlander one becomes a lighter, fully tested item with a greater output (1350Nm at the axle by my maths) also only £350 to replace when you melt something...

Either way, BMW Phev packs at 9-12kwh seem to come in around 80kg for the modules to get the full 360v at a decent current (less if going a little lower but with the outlander drive unit would be needed for higher speeds) - so probably close to 100kg with a custom case but the best way to get to 360vish currently. Add Inverter, BMS, driveshafts, mounts, ancillaries and cables and you are probably pushing nearly 200kg/20-25% increase in the car weight for the full setup so likely in my case I'd be right on top my self imposed upper limit.

[Pete9008]

Agree completely.

I started to wonder whether to ditch the Outlander box and go with the MG1 option for my current conversion but have come to the same conclusion. If the Outlander top speed is enough, and it's plenty for me on the road car, then the Outlander rear is currently the best solution.

You may be able to go with a smaller battery to save some weight, people have pulled silly currents out of Volt packs for short bursts, but I'm going to stick to the 10C limit on mine (which is still 25% over the BMW 80kW spec).

I will probably have a look at the MG1 route for the track car though. It's the only solution I have found that could cope with the higher speeds and would be quite a nice engineering project. Plus I'm lucky to have a reasonably kitted out workshop so should be able to do most of the machining/fabrication myself rather than having to pay for it (although this would be a step up in complexity compared to anything I've done before).

[Pete9008]

BTW - this might be of use, it's a spreadsheet I did to calculate weights for my conversion with the Outlander rear motor and BMW battery pack viewtopic.php?p=50527#p50527

[SuperV8]

Apologies if this is in the wrong area, if so I can move it. Long time lurker, but now starting to look at starting to build up a spec for this to hopefully build in the couple of years or so...

I have a RWD rear engined kit car, loosely based around two front rover metro subframes bolted to the front and back of a fibreglass tub (GTM K3 for those that are interested). Having a front subframe that originally had an engine in, there is plenty of space to add a small electric motor, plus the hubs are even designed to take driveshafts! The car currently only weighs about 850-900kg so want to keep as light as possible.

Something I've been thinking about doing for a long time is adding a smallish (40-70kW) as light as possible motor to the front wheels to allow more spirited corner exits, traffic light getaways or the ability to drive quietly as the 1.8l petrol engine is a touch loud for 2am village driving. I'd keep the battery small but hopefully big enough to give a few miles range if needed. Initially the idea would be to have it as a standalone/regen charging only setup but should the pack have enough capacity to be worth it (15 miles range I'd say) I'd look at adding the ability to charge from the mains.

In terms of motors I've looked at the rear drive unit from the Outlander as it's a nicely contained package that should bolt in fairly easily, but I'm struggling to find the maths on what output I would be able to expect if running at lower voltages - any pointers here?

On the subject of lower voltages, my plan would be to have a small pack for keeping the weight down but something that will allow me ability to pull high current from it for a short period - I'm not overly bothered about high speed as the existing motor will provide the power there so power dropping off above 35-45mph would be absolutely fine as long as I'm not overlooking something if I end up running the motor faster than the battery could run it? In theory the car can do 140mph so I don't want to end up with a battery on fire if I end up doing that on a track!

Also I know the lower voltages will limit my inverter choices severely, what options are there that will allow me to run lower? I can't find anything other than the volt inverter (might be wrong there too?) that will do sub 200V but I might be missing something?

Thanks in advance and sorry again if any of these are easy answers - I'm obviously googling the wrong things if so!

Ah - I nearly built a GMT K3 - in the end went with a Dax Rush V8.
Can't help with the PHEV conversion - but interested in your project.
BMW 330E PHEV has a fairly thin motor between the flywheel and autobox - could something like this be adapted to your k-series & PG1?

[maxouni]

Great idea to power the front axle of a mid engine car by electric motor.
I have a mid engine car like you, a Toyota MR2 with K24 engine and I'm going to fit Leaf motor on the front axle. As weight is a critical parameter of a performance car, battery weight kills the idea, but I think I have a solution: LiFePo4 A123 batteries with Maximum Pulse Discharge current (10 seconds): 50 C. So If buy 96 10Ah batteries, the weight of them will be about 2 kg and capable of taking 160 KW  (10 seconds). I thik 10 seconds is enough to hit speed limit. As the battery  capacity is only 3.2 KWh, plug-in charging is not important and you can charge them by recuperation only.