Lightweight AWD Sportscar

designerguy · Feb 23, 2018

A couple point for clarification:

-not in wheel motors but rather one motor for each wheel with a driveshaft...the trade off in ride and handling aren’t worth the ease of packaging

-I’m not necessarily going to reduce the rear track vs front track, however, narrowing the rear track won’t make the car inherently unstable. The rear will simply need to be balanced with the right roll stiffness vs the front to get good load transfer. Think of it as somewhere between a reverse trike and a car. The trike has no rear roll stiffness yet goes in a straight line just fine.

-longitudinal CG location affects load transfer front to read more so than straight line stability. Most stability is gained through caster and trail.

- for this car I’m aiming for as low a polar moment of inertia as possible to make the car more responsive to direction change.

Inducter · Feb 24, 2018

I see.

Notice that I said wobbly, not unstable. There's a difference between the two.Triangular setup is inherently stable, just not comfortable. That's what I was aiming at.

If you can take the cost, you go for it and enjoy. I am merely thinking aloud here. The the sheer amount of cost involved in an one-off car build intimidates me to no end. And there are technical issues, like road worthiness, comfort, safety, electrical wiring, to name a few, to consider too.

Punx0r · Feb 24, 2018

Agreed, it's a hell of a DIY project. Best of luck and please post updates!

speedmd · Feb 24, 2018

Potentially ground breaking in many aspects. Interesting trade offs on weight distribution can be explored to great extents with this technology. Spreading the battery mass out like tesla does vs a tighter central block. Inboard will certainly be lighter on its feet but possibly suffer some in ultimate traction in some conditions. Benefits may far out weigh the losses. Same with the low MI. Benefits of a car that can change direction on a dime may far out weigh any stability benefit that higher MI setup could bring to the driving experience. Loving the project.

designerguy · Feb 24, 2018

I’m trying to be really open minded to what an electric car CAN be. Many of the trade offs that IC cars make simply don’t need to exist for an electric car.

An update on the goals for the project; the Midwest and especially Michigan are having a record year for pothole damage. After driving my wife’s car with ~6 inches of suspension through some car eating potholes and nervously checking for cracked rims and damaged tires (there was some tire damage) I have decided the car needs to be not just usable in these conditions, but composed. Because of this, I am looking at ~11 inches of suspension travel for each wheel, and using a hydraulic anti roll bar for roll control, similar to creuat and kinetic. I have some different ideas on damping and mode separation so it will be interesting to try to make it work

. Luckily the car will be drivable with corner dampers and springs so that complexity doesn’t need to happen up front.

Unfortunately right now I am 100% swamped with work so not much is happening other than idle minded rumination.

Coolohm · Feb 24, 2018

This is so cool.

Why not use airbags? I designed and implemented an automatic leveling air suspension controller and system in my private passenger car and also in a large tandem horse float.

How do you plan on having the body made? I guess you have in house machines that you will utilize for making molds (assuming it's composite)?

Have you thought about using two slightly larger motors - one for front and one for rear. Will reduce both electrical and mechanical power distribution. Could use a chain / belt drive diff for the reduction to avoid gearboxes and still have a proper differential.

designerguy · Feb 25, 2018

I would like to avoid active systems as much as possible. Most tend to come with a hefty weight penalty and have a greater number of failure modes. However, I am intrigued by your system. Is it similar to an SM?

I plan on make the body plug the old school way: using Masonite sections every 4” or so and filling the rest of the body with expanding foam, then sanding, sculpting, priming, gel coating, and waxing to prepare the male plugs. For intricate sections I will probably CNC some hard foam inserts.

Differentials are great for power distribution. However, with 4 electric motors I can program different behaviors like torque vectoring and 0 point turn. This is one of the many ways electric cars can be simpler! No need for wheel speed sensors, ABS can be handled by the wheel motors to modulate each wheel in slip, traction control is simpler as well!

I don’t have much experience in these systems but my hope is that I can use this car as a test bed for open source versions. Making an electric car should be vastly simpler than a locost of similar performance. With Machine learning algorithms for suspension and chassis design (ala Hackrod) the future of DIY seems alive and well!

speedmd · Feb 25, 2018

I would keep a eye on the active suspension side as a potential option long term. Agree, that it would be much more complex and have tons of downside on the reliability side, but the potential upside could be ground breaking. Not just active, but pre-active. Gas sprung inline motor - generator actuators that could be tied into a system that reads the road ahead and picks a trajectory, where it could pre-jump humps and dips and fully avoiding those big pot holes at speed. Good to dream.

designerguy · Feb 25, 2018

speedmd said:
I would keep a eye on the active suspension side as a potential option long term. Agree, that it would be much more complex and have tons of downside on the reliability side, but the potential upside could be ground breaking. Not just active, but pre-active. Gas sprung inline motor - generator actuators that could be tied into a system that reads the road ahead and picks a trajectory, where it could pre-jump humps and dips and fully avoiding those big pot holes at speed. Good to dream.

Proactive suspension system have been around for a while. A number of them should be hitting the road in the next 5 years. It’ll be a huge boon to efficiency nuts as there won’t be energy loss due to damping.

On a side note, I have been looking more and more to move the batteries rearward behind the passengers and to move to 6 motors (2 front 4 rear) with 12 controllers. This would mean I could use 3 MTI85W100GC per controller which should be good for 18kW per controller. This would also increase the inductance each controller sees to make a more reliable system. I can move to Elise size wheels and tires since it has similar weight distribution.

Inducter · Mar 3, 2018

What is the thinking behind so many motors and twice as many controllers? Is there any advantage in it? 6 motors and 12 controllers are a lot of opportunities and space for things to go wrong and cause unwelcome costs. Plus, they'll take up the space you may want for more batteries, more power.

designerguy · Mar 3, 2018

Inducter said:
What is the thinking behind so many motors and twice as many controllers? Is there any advantage in it? 6 motors and 12 controllers are a lot of opportunities and space for things to go wrong and cause unwelcome costs. Plus, they'll take up the space you may want for more batteries, more power.

The motors are to increase the inductance each controller sees. With a 90mm stator for the same kv the inductance will be 1.5x less. It’s the same reason for the multiple controllers, higher mutual inductance for the same kv.

Another thing the multiple controllers does is increase redundancy in case of failure. Each motor will be run as 2 sets of 3 phase windings separates by 30 degrees. Basically if one controller or winding fails, the motor still works and only has a slight reduction in Km. It also reduces torque ripple.

On top of all of those things, with multiple controllers for each motor, each controller will have a simpler, cheaper powerstage. Less worry for current sharing between parallel FETs. It will also make so the power stages are relevant to this community as a high power ebike or small motorcycle controller.

designerguy · Jan 17, 2019

Just about a year on from my last post and near 0 progress!

I’ve come to some new assumptions: one is on battery voltage. Orion BMS supports LTO, many cells in series and chademo. This opens the door to quick charge but it seems that most stations do not work with packs under 200V. That means I will be connecting all the cells in series for ~350V pack.

This creates some new issues with motor control however. I was more comfortable going down the route of multiple low power controllers per motor but now I will have to build 4 motors and 4 controllers. Any ideas on 350V ~50kw motor controllers?

Another hurdle is the wind shield. Safety glass is a must in Michigan so I would have to either get a custom formed windshield($$$) or a lotus Elise windshield($$$)....So as of right now I’m assuming no windshield!

That means I will almost certainly have a higher Cd, but I will also have a lower frontal area. New challenges!

www.recumbents.com · Jan 17, 2019

Something else to think about is that most CFD systems don't take ground effects into account. I'm not a CAD expert but that's a big component of drag. It seems like most car designs leave out the aerodynamics of the underside of the vehicle.

Yes, you should be able to find a windshield from another car, and design your car around it. It's like the tail wagging the dog but will definitely make your fabrication easier. Also you could look into a pre-made tube frame chassis and hopefully find one that you can wrap that cool shape around.

Warren

designerguy · Jan 17, 2019

www.recumbents.com said:
Something else to think about is that most CFD systems don't take ground effects into account. I'm not a CAD expert but that's a big component of drag. It seems like most car designs leave out the aerodynamics of the underside of the vehicle.

Yes, you should be able to find a windshield from another car, and design your car around it. It's like the tail wagging the dog but will definitely make your fabrication easier. Also you could look into a pre-made tube frame chassis and hopefully find one that you can wrap that cool shape around.

Warren

A big part part of that is the ground/wheel interaction. You have have to model in a rotating wheel that actually crashes through the ground which is complicated to model and mesh. Most cars don’t have any underbody shielding for one reason, it costs money! ICE cars don’t get nearly the improvement from aero gains that electric cars do which is why you see electric cars focusing far more on the underbody. Even so, nice flat floors are hard to achieve with suspension clearance, service holes, stiffening ribs..... I will have a much easier time since a flat floor doesn’t really cost me anything! The only real tuning in the entrance to the underbody and the exit angle.

Xtr6 · Jan 19, 2019

Did a few years ago a build that ended with a petrol engine because of regulations in my country.
Got it road legal but it was so unpractical to drive in the city. Plan was to make it eletric but then it could only be used as racer.
Driving the car in the city Always took atraction of a lot af people.

designerguy · Jan 20, 2019

Xtr6 said:
Did a few years ago a build that ended with a petrol engine because of regulations in my country.
Got it road legal but it was so unpractical to drive in the city. Plan was to make it eletric but then it could only be used as racer.
Driving the car in the city Always took atraction of a lot af people.
IMG_20170822_101213 (640x480).jpg

Nice XTR! My hope is that the car will turn heads as much as a nice track day special

designerguy · Jan 20, 2019

So I had a talk with my wife about the car not having a windshield. It looks like It will have a windshield after all. Something about bugs and not wanting to wear a helmet every time we take a trip in the car.

On the motor side of things I might even be backing away from AWD just to simplify motor control and save some money. 350V controllers off the shelf aren’t cheap (far more than custom windshields) and the community created controllers are overkill for 50kw per motor. So I’m looking at making two motors for the rear that will spin the right and left wheels independently. Arlo’s controller and Paul and Sabrina’s controller both end up about the same size and would fit nicely vertically attached to the frame in front of the motors. I’ll most likely end up building one of the two depending on cost and my extremely limited EE knowledge.

On the chassis side I’m also looking to move to a more traditional route with two pontoons either side to house the batteries. This will allow me to move the seats closer together and reduce the frontal area of the cockpit. It will also allow me to move the batteries further rearward for a more extreme front to rear weight bias.

My computer took a massive power surge a couple of months ago and I’ve been in the process of rebuilding it. Hopefully I’ll have some CAD put together this week on some layout and design ideas.

Dauntless · Jan 20, 2019

Is that racecar an Auto Union? I'll take one.

https://www.facebook.com/LifeofaRaceCar/

liveforphysics · Jan 20, 2019

I'm excited for your project.

I think you had the right idea the first time with lower voltage. Chademo stations you find in the wild are only 50kW charging, I have 28kW of chargers fitted to my DSR motorcycle (over half a typical DC-fast charge station's power). I tap 4 * J1772 plugs, or 2 Tesla plugs when I charge as my AC source to charge my motorcycle. I don't personally use the socket, but we tested using the CHAdeMO as just a HV input bus voltage to the chargers AC input, which being as modern PFC front end electronics running equally well from DC as AC input power, it worked fine.

Anytime you have to go over 20-28cells in series, everything gets a step-function increase towards much uglier with respect to designing for life-safety, BMS cost/complexity/harnessing/failure-modes, and long-term corrosion protection if you're planning on using the vehicle in all weather and/or salted roads.

If the whole motivation for 350Vdc is just for charging, if I can solve it on a motorcycle (using 8 x DigiNow chargers), you should be able to solve packaging it in a car (maybe in a removable way if you're at the racetrack, being as each 3.3-3.6kW charger is at least 4 or 5lbs. If you're really doing the small Titanate pack, even just 21kW of chargers and 3 x J1772 plugs or a couple T-plugs (Tesla Destination AC charge stations) would let the car work fine unless your focus of the vehicle is long road trips through areas with DC-fast charge infrastructure (J-plugs, T-plugs, and NEMA 14-50 plugs at campgrounds are all across the country, often in multiples).

designerguy · Jan 20, 2019

Pretty much only for charging, yeah. 50kW is still huge at 19kwh. Most “scenic” drives leaving from southeast Michigan are 150+miles. That puts my entire range right on the bottom edge of being useful without fast charge. Stopping in a nice downtown once for half an hour halfway through the trip is nice that stopping for an hour 3 times on a longer trip. The industry is moving towards 150kW charger as the new price of entry, going up from there.

Most of the AC chargers around here are 6.6kW or less, which puts a huge emphasis on range over charge time. There are a few DC fast charge capable stations too. The coasts are way ahead when it comes to infrastructure.

I really want to use a low voltage pack, I’m just not convinced I can get the useability out of it. If chademo standards were actually followed I could go 115V and still get ~half recharge in a half hour(@125A). That would be great!

designerguy · Jan 20, 2019

liveforphysics said:
I'm excited for your project.

I think you had the right idea the first time with lower voltage. Chademo stations you find in the wild are only 50kW charging, I have 28kW of chargers fitted to my DSR motorcycle (over half a typical DC-fast charge station's power). I tap 4 * J1772 plugs, or 2 Tesla plugs when I charge as my AC source to charge my motorcycle. I don't personally use the socket, but we tested using the CHAdeMO as just a HV input bus voltage to the chargers AC input, which being as modern PFC front end electronics running equally well from DC as AC input power, it worked fine.

Anytime you have to go over 20-28cells in series, everything gets a step-function increase towards much uglier with respect to designing for life-safety, BMS cost/complexity/harnessing/failure-modes, and long-term corrosion protection if you're planning on using the vehicle in all weather and/or salted roads.

If the whole motivation for 350Vdc is just for charging, if I can solve it on a motorcycle (using 8 x DigiNow chargers), you should be able to solve packaging it in a car (maybe in a removable way if you're at the racetrack, being as each 3.3-3.6kW charger is at least 4 or 5lbs. If you're really doing the small Titanate pack, even just 21kW of chargers and 3 x J1772 plugs or a couple T-plugs (Tesla Destination AC charge stations) would let the car work fine unless your focus of the vehicle is long road trips through areas with DC-fast charge infrastructure (J-plugs, T-plugs, and NEMA 14-50 plugs at campgrounds are all across the country, often in multiples).

Re reading your post multiple times, a few questions.

How is charging on destination chargers? For higher rates are you using multiple chargers?

How was chademo handshake handled? At what rate was it able to charge?

I just see chademo and ccs taking off. VW is building chargers that support both standards as part of their settlement, Hyundai/Kia have adopted chademo...

LeftieBiker · Jan 21, 2019

I just see chademo and ccs taking off. VW is building chargers that support both standards as part of their settlement, Hyundai/Kia have adopted chademo...

In the USA, CCS is the better bet, as combo stations are coming online with several SAE plugs and just one Chademo plug. Aside from Nissan, manufacturers seem to be hedging their bets and building cars with whatever charge port is common in the country of sale. Even Nissan is rumored to be looking at switching to CCS in Europe...

atarijedi · Mar 4, 2019

Will you be driving this in the winter? Because it seems kinda low for snow, or speed bumps, lol.

Anyways, for the motors, you should check to see if Revolt can do Halbach Arrays with the magnets. It should increase the efficiency of the motor, and higher power density. Might cost a bit more, because construction is more difficult, something to ask them about though.

Lightweight AWD Sportscar

10 mW

100 µW

1 GW

10 MW

10 mW

1 mW

10 mW

10 MW

10 mW

100 µW

10 mW

10 mW

10 kW

10 mW

100 W

10 mW

10 mW

100 TW

100 TW

10 mW

10 mW

100 kW

100 W

Similar threads