How far do you want it to go?
Almost every discussion of electric vehicles STARTS with the concept of range. This is because it has always been the defining limitation of electric cars. In the early days of the Internet, electronic bulletin boards, and online communications, it was always “how fast is your modem connection.”
Fast is relative. Incremental advances are NOT all created equal. The jump from 300 baud to 1200 baud was an ENORMOUS improvement. The jump from 1200 to 2400 was great. The jump from 2400 to 9600 was a good improvement. To 28.8 kbps was good too. And to 56.7k was a bit of a yawner.
Range similarly. The difference between 25 mile range (and falling) provided by Pb chemistry cells and the 80-100 mile range offered by LiFePo4 cells is ENORMOUS. From 80 to 200 would of course be very good. After that, I doubt that it matters to most.
Indeed, the average daily use of automobiles in the United States is 39.4 miles for about 14,500 miles per year. And over half the driving population averages under 26 miles per day. People just don’t drive as much as they think they drive. But of course, owning a car means being ABLE to go as far as you like, not so much that you DO.
The internal combustion engine driven vehicles of course vary in range depending on how you drive the car. And they are about 12% efficient at translating some 33kWh of energy in each gallon of gasoline into forward motion. By far the majority of it is converted to HEAT and blown out through the radiator in the front of the car – each vehicle contributing to our warmth and comfort during the winter.
Electric vehicles are more like 85% efficient. But that efficiency is precisely what makes them MORE sensitive to just how they are driven.
In this weeks show, we begin a series of tests to demonstrate the factors in the vehicle design that affect this range for a given driving scenario. And we compare two cars that SHOULD have the same range, and somewhat mysteriously do not.
[jwplayer file=”news081911 – iPhone.mov” hd.file=”news081911-1280.mov” image=”http://media2.ev-tv.me/news081911.jpg” streamer=”rtmp://s2v8uso6bi7t47.cloudfront.net/cfx/st” provider=”rtmp” html5_file=”http://media2.ev-tv.me/news081911 – iPhone.mov” download_file=”http://media2.ev-tv.me/news081911-1280.mov”]
In my own opinion, we need to strongly alter our relationship to our vehicles and how we view them and relate to them. I do NOT actually fall into the camp where we should “suffer” for the sake of the planet. Technical problems usually can be attacked with technical solutions and this one is no different. Keep your Yukon or Denali for towing the boat to the lake or the trip to the nearby city to pickup someone at the airport. Take your long distance trips across the country as much as you like in your Escalade or Suburban. You aren’t hurting ANYTHING because you are not doing enough of all that to move the decimal place. If typical, 95% of your driving is to the hardware store, the lawn center, the grocery store, the soccer field, and to school and work. And the mileage you are ACTUALLY driving in any one gulp is trivial – in some cases within Pb chemistry range.
But cars have become more expensive. In the 1960’s, $4000 was a LOT for a car – picture a new Vette or a Cadillac. Today, $50-$80K is NOT stratospheric. That’s the cost of a house in years past.
Now you wouldn’t send a house to the junkyard after 9.6 years and build a new one. You might remodel the kitchen. You might do a little paint and carpet. But the very deliberately planned obsolescence from our friends at the automobile factory has led us to some ridiculous concepts. Today, you can TOTALLY REPLACE the drive train in any automobile – engine, transmission, and all belts and hoses, for LESS money than the SALES TAX is typically on a new vehicle. You COULD put an entirely new interior in a car for about $3000 – done by professionals. A paint job? if you want to be the rage at the next car show- $6000-$7000. But Earl Scheib DID make a fortune with his “ANY CAR – $49.95” paint job.
And so my intent is to pick cars I really like, and then just renovate them as needed. That pretty much precludes the use of steel bodies. I like aluminum and fiberglass because they don’t try to become biodegradable and return to the earth as rust while I drive them.
With the advent of these LiFePo4 cells and electric drive trains, I can see simply continually upgrading a car forever. We’ll not build it. We’ll build it for life and we’ll keep building it. As batteries become available, or controllers, or chargers, we will upgrade. If the paint fades and accumulates some scratches, we’ll have it painted.
Back to the main topic of this week’s show. Range. It does require the application of power to move a car at all. How far it moves is a function of how much power. There are several factors that determine just how MUCH power is required. Since your battery pack offers a FIXED amount of power, your range is the result of your relationship to those factors.
INERTIA. The tendency of a body at rest is to remain at rest and for a body in motion to remain in motion. Without other factors, if you had a car going 25 miles per hour, it would go 25 miles per hour forever with NO additional power. Friction, and other factors cause it to decelerate – eventually to a stop.
You have to apply power to accelerate to any speed, and other resistances will cause you to DECLERATE from that speed to zero. The amount of power required is a function of MASS. The greater the mass the more power to accelerate.
This is entirely separate from gravity. But when we combine gravity and mass, we can derive a very measurable quantity – WEIGHT. And so the amount of power required to accelerate to a given speed can be though of primarily as a function of weight.
GRAVITY is indeed one of the factors in electric car design. All roads have some incline. To climb a hill, you must move the mass in elevation against the force of gravity. In a downhill run, gravity can actually decrease the amount of power required to a negative number, and indeed you can recover energy through regenerative braking – converting both inertia and gravity into electrical power to recharge your cells.
AERODYNAMIC DRAG is actually one the most significant forces involved in range. As anyone who has put their hand out the window at 60 mph knows, the force of the relative wind can be felt as quite a strong force. That is aerodynamic drag experienced across probably 20 square inches of hand. In the case of the Spyder or the Speedster, that same force is multiplied by the 1.76 square meters of frontal area if you can imagine it. It is an enormous force.
The aerodynamic drag force can actually be given by the formula Fd=1/2 pv2CdA.
Fd is the resulting drag force.
p is the density of the fluid – in this case air and is slightly variable by temperature.
Cd is a “drag coefficient” typically derived from wind tunnel testing to account for the fact that for a given frontal area the drag will vary according to the shape of the vehicle and the resulting laminar flow of air across the body.
A is the total “frontal” area of the car. How big is the hand.
v is the culprit – velocity. It is squared. In this equation, whatever v is is multiplied by itself. If it is 10 mph, it equals 100. And if it is 100, it equals 10,000. This cruel fact is what caused such an enormous difference in power at 70 mph than at 40. Recall Jimmy Carter’s 55 mph speed limit. It was not invoked to save lives – but rather to save gasoline. And as I mentioned, electric cars are MORE sensitive to such things than the gasoline cars because they are MORE efficient in the first place.
After inertia, gravity, and aerodynamic drag, we have rolling resistance. If you put your car in neutral, and push it out of the garage by hand, you are experiencing it’s rolling resistance first hand.
The biggest component of rolling resistance is probably the tires. Large fat squishy tires give us a very comfortable ride. But the flexing of the sidewalls to accomplish this generates heat loss and makes the car difficult to roll. The larger the contact area of the tires with the pavement, the larger the rolling resistance. If you don’t have pavement, and are driving on sand for example, it gets worse.
But there are drive train resistances as well. The friction of the gears in the transmission and differential add a bit of drag to the system. The bearings in the car support the entire weight of the car on the wheels while allowing wheel rotation. Whatever your car weighs, each wheel has to carry part of that and it does so primarily through the wheel bearings. They have friction, and give off heat. If you have one that goes “dry” and is not lubricated, it will generate enough heat to melt steel.
This week, we did some testing of the 1957 Porsche Speedster Electric we call DUH, because it was our second speedster build – part deux. And we also tested a 1955 Porsche 550 Spyder electric replica that we consulted on when it was originally built by Duane Ball and Scott Smith, and subsequently we bought the car.
We tried to eliminate INERTIA and GRAVITY from the test. We did this by driving a substantial distance of about 20 miles at a steady speed. Not accelerating and not decelerating pretty much eliminated the application of power for the purposes of inertia – just what we needed to get up to speed in a few hundred yards out of the 20 miles. And even that was largely cancelled by the recovered energy or lack of force required for the distance we decelerated.
All roads have slope. We divided the 20 miles into two halves and did them in opposite directions on the same road. This should pretty much cancel the effects of gravity.
Leaving aerodynamic drag and rolling resistance. We tested both cars on the same route and nominally at the same steady speeds of 40 mph, 50 mph, 60 mph and 70 mph. Let’s look at the results.
If we look at these results, a couple of things jump out at you. At 40 mph, the Speedster gets 19.69% greater range at 158 miles than the Spyder at 132 miles. Also curious is that this advantage decreases to 9.3% at 70mph with the Speedster coming out to 94 miles and the Spyder at 86.
That brings up two questions: Why does the Speedster have greater range? And why does this disparity vary with speed?
And these questions jump from curious to bizarre when we look at the two cars.
Coefficient of drag:
Spyder : 0.20
The Speedster should be LESS efficient than the Spyder by quite a bit. I confess the references to Cd on the Spyder are slim. There were only 90 of these vehicles built originally and wind tunnel testing wasn’t a big deal in those days. But just looking at the two shapes, you would expect the Spyder to be dramatically more slippery through the air. Both vehicles have a frontal area of 1.76 square meters.
As we took the weight largely out of the picture by eliminating Inertia and Gravity, that’s not a big mystery.
The cars are similar in other ways as well. They both have 36 China Aviation Lithium Battery Company cells of 180Ah per cell. They operate at the same voltage. They have the same High Performance Electric Vehicle Systems AC-50 motor. They both used the Curtis 1238 three-phase controller. In this, the Spyder features the 650 amp -7601 controller variant while the Speedster has the 550 ampere -7501 controller. But we never went over 200 amps in any event. Both controllers were set for 5% neutral braking and 500 rpm taper. These are controller configuration items controlling the amount of regenerative braking.
They both now use the same hall effect accelerator.
They DO have different transmissions with the Spyder having a 3.88:1 ring and pinion and the Speedster featuring a 3.44:1 ring and pinion. But the other gears result in almost identical final ratios between the two cars. Alll the testing was done in fourth gear. There was a 100 rpm difference at 60 and 70 mph but almost identical rpm at 40 and 50 mph.
The Spyder features a GPS speedometer and the Speedster features a cable driven speedometer calibrated in MPH. Or should I say miscalibrated. There is a 2.3 % difference from the GPS which we assume to be true distance and speed. This is undoubtedly part of the mystery, but doesn’t quite cover the question in degree I’m afraid. It is less than 2 mph and that kind of gets down into the noise level of what I can maintain as far as a steady speed on an Interstate highway with traffic. We used the distance measured by the Spyder for all calculations.
The clue is in the disparity between the 40 mph results and the 70 mph results. If aerodynamics accounts for 1/2 the difference between the two, and rolling resistance the other half, as we increase in speed the aerodynamics should increase as a percentage of the total as aerodynamic drag increases. Rolling resistance may increase, but not as a square function and really not by much.
In this case, the difference between the two vehicles decreases from 19% at 40 mph to 9% at 70 mph. What this tells me is that as aerodynamic forces become a LARGER percentage of the total problem, the difference between the two vehicles DIMINISHES. And so the difference is NOT attributable to aerodynamic difference at all. It must therefore be attributable to rolling resistance.
The Spyder has a set of 165/65 R15 Khumo tires on it. The Speedster features a new set of Michelin Energy Saver All Season 185/60 R15s. It also has some Moon Eye wheel covers which might lend a bit of aerodynamic assistance.
The Spyder started life as a WIDE FIVE option from Beck Speedsters. This is historically accurate but problematical. There aren’t many wheels drilled to FIT the Porsche wide five pattern. So Duane had put a set of 911 aluminum wheels on the car with an adapter that moved the wheels out about an inch. This caused the wheels to rub on the body fender in front when steering. So they jacked up the front of the car – probably hosing up the aerodynamics in the process.
In any event, I never liked the setup in the first place. And there were “noises” from the rear of the car. I kept telling Brian I had a brake dragging but he never could really find any evidence of that.
In any event, we contacted AIRKEWELD. They make a very interesting aluminum brake rotor that has a billet aluminum hat that bolts on – allowing them to do any wheel pattern and offset desired with a very lightweight rotor. They couple that with a Wildewood aluminum caliper. This little puppy would reduce our unsprung weight by about 18 lbs. Aluminum rotors and calipers are not nearly as durable as steel and we would not normally recommend their use on a car. But this car features regenerative braking anyway. We don’t really use the disk brakes much at all. So we think it would be a good upgrade for this particular electric car.
We add a lightweight Weld aluminum wheel in a Ford 5 by 4.5 pattern. And on this, we’ll mount a low rolling resistance tire. The Michelin Energy Saver A/S curiously is SOLD OUT NATIONALLY. You can’t get em – anywhere. Brian has confirmed with Michelin directly that they do not exist at the moment. They are a little mysterious as to why this might be. We suspect they are redesigning for even lower resistance.
Tirerack.com tested low rolling resistance tires on a Prius and Michelins got the highest mileage at 53.8 mpg, compared to the Goodyear Insight that comes standard on the Prius at about 51.5 mpg. But the Bridgestone Ecopia rated just a little behind the Michelin and is $25 per tire less AND they are not unobtanium.
The Bridgestone and Weld combination is about 2 lbs lighter than the earlier Porsche style aluminum wheel and Khumo. The Bridgestone Tire weighs 17 lbs and the wheel a little over 12 lbs.
Unfortunately, AIRKEWELD is an American Company. ANd like all non-Chinese American companies, they cannot manage to get an order in a box and shipped from Arizona to Missouri without at least three tries. We got the long axle version on the rear hubs when we had specified the short axle. We got the wrong brackets for the front. And we go the wrong brake lines for the rear. ANd we got the wrong brake pads for the Wildewood calipers in the front. It has been a total comedy and we have now received our THIRD shipment from these guys. The parts were ordered in June and received in late August (on three different dates of course). And they were $1800, which approaches 10% of the cost of the original vehicle roller.
But they were very nice and the units are just gorgeous. We should have them on this week, and be ready for retesting. It will be very interesting to see the results.
We are also installing some Silicon Nitride ceramic bearings on the rear axles of the Spyder. This is a standard 6306 bearing which normally costs $12.95. The ceramic version is $325 each. But they are good up to 2550F and have significantly lower friction. I guess I think we are stringing the bounds of reality here at a $650 upgrade to avoid bearing friction. But the Illuminati team claim it is their secret sauce on their vehicle which gets very good efficiency on the highway and is nearly 3000 lbs to boot.
We don’t do very well with secrets at EVTV. If you have any, tell them to someone else if you want them to remain a secret.
68 thoughts on “How Far Does Your Electric Car Go?”
Did you think to check the bearing and rotor temps of the Spyder after a drive? That should have told you if you had a dragging brake or a stiff bearing. Would also have been interesting to compare to the Speedster temps. You might compare the two vehicle temps with the new ceramics in the Spyder.
I wonder how accurate the 0.20cd figure is for the Spyder. That seems really low, especially for an open cockpit. 0.20 puts it in the range of some fairly radical concept vehicles, none of them ragtops.
Any differences in wheel alingment between the two (camber, toe-in)? Nothing like scrubbing off rubber to kill your mileage.
Each wheel bearing carries the vertical load of its share of the vehicle weight, and the horizontal load of that wheel’s rolling resistance. Bearings for the driven wheels additionally carry the acceleration and wind resistance forces horizontally. The rear (driven) wheel bearing loads are higher (assuming relatively even weight distribution), but enough to warrant special bearings? If your “thrust to weight” limit on your tires is 1.0, and your weight is evenly distributed, your bearing load in the driven wheels is, at most 42% higher than the non-driven wheel bearing loads, just before your tires start spinning. At steady 70 mph, its closer to half that difference. Maybe upgrading all four wheels is in order.
There’s a company called MicroBlue Bearings that adds a coating to their ceramic bearings that supposedly decreases friction even more than ceramic alone (http://www.microbluebearings.com/). I’ve seen some impressive videos on YouTube using these bearings. One in particular of a motorcycle wheel free spinning for over 3 minutes.
I’d like to see how they fair under EVTV’s legendary scrutiny.
My suggestion, before tearing things apart, would be to take both cars out and drive them around to maximize the consumption differences. When you and your other driver get back to the garage break out some infrared thermometers and aim them at all kinds of different points on both cars. Check motors, bearings, transaxles, anything that you see that moves or covers something that moves. It might turn up some places to look.
0.38KWH difference at 40mph. It’s a lot.
I’m not going to play the know-it-all here so please take this as conversational.
I would of been tempted to do torque-to-rotate tests through the back wheels (one raised) while in gear. Both cars, before and after modification. To see if something in the drive train is awry.
If one moves the graph lines for same consumption at 40mph then the Spyder might have 9% saving over deux at > 100miles @ 70mph?
Considered a body flush rear wheel cover on the Speedster to possibly increase speed/range? Only saying Speedster because it should look good and authentic.
Actually rear wheel covers were also incorporated in some of the later model 550 Spyders. I’ve thought of this. The EV1 and the Honda Insight both use these so apparently they are effective.
I expect the changes to help, maybe even a lot.
but to get a really efficient car it has to be done from scratch because cars have been made with great indifference and there is only so much that can be done to retrofit. every aspect of a car needs rethinking. you have to imagine a car at maybe 400kg and a Cd below 0.2 and then make the parts according to that, not generic parts meant to carry 1400kg. it’s brutal but it’s the truth.
one semi easy possibility would be to make a fiber glass copy of the EV1 body work to get decent aero without too much work. fiber glass chassis too, super thin wheels, light motorcycle disc brakes, 100kg battery, no gearbox. etc
everything becomes lighter when you think light. even much cheaper although the first will be custom.
it’s also possible to make a 97% efficient motor.
the potential for improvement compared to a normal car is great.
smart is of course smart. always was, always will be
I had a 10% range gain after a front end alignment. Less road noise now as well. Maybe that will offset my A/C addition this winter.
What you have been saying here sounds soo much like what I’m in the process of making. But its small and if I do make it dual seat, she’s tucked tightly in behind with her legs akimbo around me to keep that pesky CDa down and CG good. Might settle for the 10~20KW BLDC Golden motor and controller instead of an Agni.
However, what I’ve not done is go for new components on the running gear. If I did, all my parts would be from grass track racing car suppliers which are super light and made for a relatively fair price. Sorry its a Brit. site but here’s an example page of their engineered items:
Will someone hide Jacks chequebook a minute please? 😉
I hear what you’re saying. A built-from-scratch car, designed from the ground up to be a super-light and super-aerodynamic electric car is clearly the way to go. Trouble is, you’ll be spending 80% of your time building a car, and 20% making it electric. Umm, maybe 90/10. Do you really want to reinvent door latches (how DO you spread the point load of a door latch over a thin carbon fiber panel?), or do you want to have an electric car you could be driving in 6 months? Don’t get me wrong, my heart is with you 100%. Every time I think about building an electric car my mind immediately turns to pulling molds off some exotic sports car and laying up my own carbon fiber chassis. Will I do it? Almost certainly not, my head won’t let me. It’s too much work, and time, that’s only peripherally related to the ultimate goal.
On the other hand, I don’t know you and it’s entirely possible you’ve been building race cars or hot rods for the last 25 years and you think the car building is the easy part of the project. If so, GO FOR IT! A carbon fiber EV-1 replica? Oh yeah, make one for me too. What museum will let you pull molds?
Jim, for fear of controversy and directing Jacks own blog in another direction (again).
Working around existing bodywork on a vehicle made for a different set up is not always straightforward. Jacks ‘Smart car’ is a good example. The kit car route is a good way to go and it seems Special Editions are coming on board with a killer kit. We end up with a ‘sporty special’ with good weight characteristics and not too many workarounds.
On the other hand check out the X-prize winners like the “Illuminati 7” (For build time).
One X-Prize production vehicle; is a top money all enclosed CF bodied motorcycle.
An example that I’ve mentioned before. Have recently been helping a friend on his Berlingo Electrique (a ready made electric van, modified, 600AH @ 167V) back on the road after the old cells died. Into a powered roller. *Three years* of spare time because of modifications and workarounds. On the home run now….
Silently floating down that road felt soo good!
Regarding your discussion on gearing, if you stay in one gear (here it’s fourth), mph and rpm must have a linear relationship. Plotting your numbers, the Spyder shows this but the Speedster does not quite. Given that you used the distance measurements of the Spyder for both cars, perhaps it’s that you couldn’t precisely maintain your speed or didn’t turn at precisely the same point. This small error shouldn’t diminish your analysis or affect your Spyder to Spyder comparison; I just wanted to make this point.
BTW, when weighing your disks/hats you measured 14.8 for the new and 23.7 for the old and came up with a 9.9 pound difference. Brian passed through the correct answer on the way to the wrong one at 8.9 pounds. Using the scale’s readout, I added up 59.2 pounds of total weight savings including the wheels.
Dan always has this mantra of weight and aerodynamics. He’s never built ANY electric car and only recently ridden in one.
That said, weight and aerodynamics certainly count in an electric car. Aptera stands in my mind as the preeminent example of this gone wrong. They got the weight down and had the aerodynamics, but when presenting to potential investors at a critical point in their development, one of the audience asked how they roll down the window. The CEO went into some length about the aerodynamics and the advanced environmental system that made rolling down the window unnecessary. The follow on question “then how do I order at McDonalds.”
The automobile has evolved over the last 125 years from Karl Benz’s internal combustion tricycle. The door latches have moved around. The pedals have moved around. And today a car is what we believe a car to be. Radical departures from that are faced with having to reinvent 100 years of experience on a blank sheet of paper. I’m not up to that. Perhaps Dan is. So far he has designed nothing.
But we can make improvements without throwing the baby out with the bathwater. I like aluminum as I have a good deal of experience with it from aviation. It is not true that it does not corrode, but it doesn’t INSTANTLY start trying to turn to dirt like steel does. I’m entranced with fiberglass and carbon fiber. But even Tesla shied away from carbon for a production car. It takes too long to manufacture and costs too much. Maybe not a dissuader for us at the onesey twosey level though.
As to weight, we struggle this week. We’ve had at least two MORE shipments from Arizona. This guy CANNOT get an order right. We now have two LEFT calipers. But we have done some more careful weighing.
In the front we lose 20 lbs total per side, reducing the unsprung weight from 69.4 to 49.4 lbs. This includes caliper, bracket, rotor/hub, wheel and tire.
In the rear, as we use essentially the same style caliper to keep our parking brake, it is more like 10 lbs per side. And so we are seeing a 60lb weight saving total.
But just the feel of the wheels turning in my hand, I think we’ve made some enormous gains in rolling resistance.
THe future? We’re hoping for a complete redesign of the Speedster chassis, and an option for all aluminum. If we couple that with these running gear changes, I’m confident we will have a dramatically improved car – in an incremental sense.
After that, we can take a closer look at what it would take to lighten and strengthen the body, while perhaps upgrading the paint/finishh level at the same time.
And of course along the way we’ll be looking for ways to improve our drive train – 200Ah Sinopoly’s for example. Perhaps a 145v controller.
I can tell you from the test drives that Speedster Duh is just hands down the best build we’ve ever done. The more I drive this car, the better I like it. It’s extremely agile, soft in the paws and alert in all regimes. Very nicely balanced. I do not find the same characteristics in the much more powerful 150 mile range Redux.
Continuing in that thought line, it occurs to me that focusing on the objections of the body politic in car design is not an optimal approach. The fixation with range is more properly addressed with better batteries. But in the meantime, I had hoped with this upgrade to illustrate some of the real components of range and how very variable it is.
The 8x efficiency gain is a double edged sword. It makes differences in weight and aerodynamics MORE apparent and so more critical. it exacerbates differences in rolling resistance. And it magnifies differences in driving style.
So to my way of thinking, this continues a Darwinian evolution in the automobile, this time in a more productive direction of electric drive. But it is not so much a sudden departure as a continuation with slightly different criteria. The child mind concept of “lets’ start over with a blank sheet of paper and do it right” is a recipe for disaster. It MIGHT be an EDUCATIONAL recipe, but the outcome would not be good.
That being the case, we should probably be working with the latest Porsche top of the line designs. I find it easier and more appealing to me personally to return to the Porsche root with the 356 replicas. They are relatively inexpensive, and of course eschew all the more or less necessary complicating elements such as air conditioning, heat, and satellite radio. There is no ABS or variable steering. And that lets me focus on the electric drive part.
But the shape of the car is iconic. That is, it is not subject to the whim of style and fad and season. We have little kids in the neighborhood who think this is the coolest car on the planet. They don’t know its a Porsche and they don’t know it is from the 1950s. They have no frame of reference at all.
But I could introduce this exact shape in 2050 as an automobile and it would be attractive again. I personally think I could have people standing on the HOOD of a Tesla Roadster to get a better look at the Porsche 356 shape – even when it is clearly NOT even a true Porsche.
BTW. It looks like we will have a vendor display area at EVCCON 2011 with about 20 booths. And Special Editions is bringing a roller to display.
Andy, do you have pictures of your design? I did this design for a tandem seater http://www.zev.dk/
Jim, I’m not advocating everyone build from scratch. I’m talking about what would change the world.
Jack, I understand you notion of evolution but it’s just not true for where we are going. car makers will say the same; let’s not do this whole EV thing, let’s just improve the mileage by a few miles per gallon per decade and let’s look at it again next century.
you don’t even have to imagine a lot, your kit cars are already a good step of the way. their construction is quite different from normal cars yet they work just fine don’t they.
they are incidentally fiber glass, it just has to be a bit different because they never cared about weight but fiber glass is a very strong material. good for lightness.
and you’ve remarked the chassis is heavy. a tube chassis can be quite light or a sandwich fiber glass chassis.
then just imagine 70mm wide normal wheels but thinner material because they need to carry less. tiny disc brakes for the same reason. tiny hub, small shocks etc.
try also to resist the temptation to pile on batteries. remember light begets light. you actually need proportionally less batteries than before. preferably much less. say 120V of 90Ah. or preferably A123 20Ah cells, say 120V 100Ah which is under 90kg.
you want it to move at the speed of light. that’s where the magic is.
Hi Dan, For an idea of what I’m making, Google “MEV etrike”. However, it’s similar, not a copy.
You will have to read up on what you are allowed on the road under EU rules. The rest I will mail you on issues.
You will hate me 🙁
One last word on carbon fibre. It’s a health hazard that is at least every bit as bad as asbestos. Not what you want to breath in after a smash and for the greenies; not recyclable.
I don’t hate you. I think you are wrong on one key aspect of EU law though. most other points I have already considered.
yes carbon fiber appears to be quite unhealthy. the easy fix is simply to use fiber glass which is a lot cheaper too. and even if you’d use carbon, the amount of exposure in the rare crash has to be insignificant. it’s more a concern for the people who work with it but that can be handled with bunny suits or by machines.
will you make your trike open like that or give it aero bodywork?
Jack, Firstly, I am a little confused about the idea that you can eliminate gravity as a variable simply by going both ways along a fixed route to establish the efficiency of your cars – unless you are only comparing one vehicle with the other (which probably is what you are doing). To arrive at an absolute Wh/mile, the only way to do it would be to use a level stretch of road… but, as I say, I expect you know that.
As far as the curious case of the Spyder’s lower drag not equating to lower efficiency, it would be interesting to see what the theoretical forces actually work out at for both cars (in N) and to compare those related to aerodynamic drag to those related to friction. I can’t believe that at anything more than 30mph or so the latter are any more than 5% or so of the former. If so, something cosmic is going wrong. To find out what (without spending a fortune on whacky bearings and the like) would it not be an idea to do an extended rolling road test on both vehicles measuring the power in & out as this would effectively eliminate the aerodynamics almost completely? Keep up the good work. Regards, MW.
One step at a time Dan 😉
We can only speculate the outcome of the Spyder. Can’t wait for next Fridays nail biting show.
Andy, I don’t speculate, I deduce 🙂
the Cd of the 550 is not 0.2. much closer to the other one. maybe 0.37. difference has to be drivetrain and I’m guessing gearbox (thick oil maybe) and tires
verb (used with object), -duced, -duc·ing.
to derive as a conclusion from something known or assumed
1. Taken up or used so as to deceive; pretended:
2. Taken for granted; supposed:
1. Presumed to be true or real without conclusive evidence.
which leads to:
A conclusion, opinion, or theory reached by conjecture.
Reasoning based on inconclusive evidence; conjecture or supposition.
1. Inference or judgment based on inconclusive or incomplete evidence; guesswork.
2. A statement, opinion, or conclusion based on guesswork:
v. con·jec·tured, con·jec·tur·ing, con·jec·tures
To infer from inconclusive evidence; guess.
In the end its all a GUESS.
Ooops. Forgot this one.
1. The act of supposing.
2. Something supposed; an assumption
GreenEV, it might look like guesswork to you but I’m not guessing when I say the 550 Cd is not 0.2.
I can tell from looking at it and I can tell from the data he got.
the 45 degree windshield ending in nothing is a cardinal sin in aerodynamics. when glider planes do that with a small fraction of their wing their total drag quadruples. the whole aircraft. that’s how bad that is.
and the data shows the curves growing almost identically. because aerodrag grows increasingly with speed and rolling drag is only proportional you can deduce which part is aero and which part is rolling. because they curve up similarly their aerodrag is quite similar so there can’t be that factor 2 ratio in their airdrags.
I’m not sure who the source is for the 0.2 claim but they are making it up. if you remove the windshield and cover the cabin with a smooth plate it might get 0.2 but that’s something else.
I’m sort of guessing that it’s the transmission oil and the tires but not really. the mechanical noise could be a wildcard and there could be other unknowns. maybe bad bearings in the transmission etc
An open cockpit is bad only if the air pressure in front and behind is unequal. Open cockpit gliders (Which I have flown) do not exhibit losses because the air smoothly recollects around the headrest/fuselage. They are very quiet to fly in.
I keep on giving references to Craig Vetters website/motorbike because it is exactly the same in function but nobody is listening. In fact the sides are wide open too. No turbulence therefore show no loss of mpg or speed.
Jacks Spyder has a lower CD than Deux. The proof is in the graphs where Deux gains 9% losses at speed against the Spyder between 40mph and 70mph.
The graphs suggest that the Spyder is slightly more aerodynamic than the Speedster. It doesn’t look like a major issue as the lines point together pretty gently. The 3.44 R&P is, in theory, more efficient than a 3.88 R&P but the difference is small enough that I can’t imagine that size effect. That leaves unaccounted for mechanical drag and tires. I can’t see Jack missing a damaged gearbox or wheel bearing.
I’m inclined to agree with Jack. I think this is mostly tyres (the Spyder tyres — I couldn’t resist.) That is eye opening.
Andy, I’ve looked at the site. it’s quite different to have a gap in a continuous surface compared to just a wide 45 degree up windshield.
and even a gap will cost you.
it would be an interesting experiment to put a thermoplast bubble on top of the spyder instead of the windshield. combined with a thermoplast plate to make a smooth underside the difference will be quite dramatic
If you KNOW and are not guessing, and I happen to know this is more of your fantasy world actually, then please provide a link.
I found one indicating 0.20 and I too look askance at that. But the way you would know is from a link to a known value derived from wind tunnel testing. What you can “see” and just “know” is not very interesting.
That’s kind of the difference between the old bull and the young bull.
I’ve “known” so many things that were wrong that I no longer place much value in what I “know”. Somewhat more in what I can repeatedly demonstrate.
Just finished editing this weeks show. It was not as conclusive as I had hoped. We’re kind of down to the transmission, and have actually kind of taken most of the gearing of that off the table. Mayhaps an alignment issue somewhere, but it does’ quite jump out and a small one again wouldn’t quite cover it I would not think.
It’s a mystery.
Jack, you’re obviously familiar with the scientific method. What you’re doing is advocating evidence, not what you think you know. But you are still willing to give an opinion, and “fall on your sword” if you’re wrong. That’s quite admirable. OK, enough kissing arse.
It would be nice to have a car that would go as far possibly, as efficiently as possible. Another way I’d like to beat whatever arbitrary range barrier we can think of is fast charging like the leaf. Is that possible for us now? Or can you talk about it?
I test drove the Leaf recently, it was great. If I bought it I know I could drive to Galway city from where I live (I live 100 km from Galway), but I’m sure I’m not going to make it back without charging. But as luck has it the Nissan garage in Galway has a fast charging station (for free), so now I know I will make it back. For this reason I think the cars converted in the future should have the facility to fast charge – if possible. I “know” Jack you are happy pottering around to and from the shops, but I’m not. I also suspect that that 5% of driving is quite important to people and perhaps feels as if were 20% of their driving. I’d love to hear you opinion.
P.S. where’s the video?
Rear wheel alignment.
Tried an in-gear pull test on both cars in the garage with scales?
If this does not show anything conclusive I would suggest an exorcist.
Yes, I’m advocating evidence OVER what you think you know. This because what I think I know has proven unreliable over the years. I’m aware of it. THe problem is that most people are not. They are unshakably confident in what they know, without any evidence at all. So for me, such conversations are a little constrained, – kind of like describing the intricacies of your marital relationship to your cat.
The cat seems attentive, and even a bit smug. But it’s a cat.
Your reaction to the Leaf is quite legitimate and your desire to travel to Galway as well. We have St. Louis. 110 miles north. I have other vehicles for that, but yes, certainly that sort of travel would extend the usefulness of our electric cars.
Fast charging is obviously the solution. Do we have that capability? No, but it is tantalizingly close. The batteries are well able to charge in 20 minutes. In fact, I routinely charge 100 Ah or 90 Ah cells at 200 amps in the lab. They don’t really even warm up. It has become VERY routine. This would charge a car in about 30 to 45 minutes depending on your definition of charged. It would be well over 90% charged in 30 minutes. The last 10% is a bit problematical.
Other than lacking a standard and a 3 phase power hookup, the EVSE is not really a problem either. A standard connector is due from SAE Q1 2012.
The charge process itself is the problem. Level III charging, much like Level II, is not really managed by the EVSE at all. It simply makes a high voltage (400v?) high current (125-300 ampere) and thus high power (up to 120kw) source available to the car. Just as your charger feeds off the Level II EVSE for 240 vac, and actually manages the entire battery charge process internally, a Level III charge requires an onboard charger that can accept the 400v and 300 amps and make use of it.
The CC/CV charge curve still has to be managed by the onboard charger in the vehicle. Only in this case it has to be managed at 300A.
Actually, the charger you have now can deal with the 400v dc. Inside your onboard charger, a bit of circuity is devoted to converting the 240 AC source into about a 310vdc source – a rectifier.
But the onboard chargers typically do 15 to 20 amps output to the batteries.
Now where in our electric car could we get a device that can manage 300 AMPERES of power to the battery cells?
Well, the DC CHOPPER CONTROLLER or PWM CONTROLLER actually does a very controlled very of precisely the same thing for the motor. It takes raw DC voltage and current from the battery pack and applies it in very controlled fashion to the motor in proportion to your throttle inputs.
What would it take to convert a DC CONTROLLER to a FAST CHARGER? Not very much actually. Some software for the most part. The controller can already measure pack voltage. And it can already throttle voltage and current. And it can already accept in some cases up to 360v of input voltage at up to 1000 amps.
If you turned the controller around, and connected its’ output to your batteries, and its input to the fast charger, and used the throttle to control the current into the cells, and a voltmeter to observe the rise to your CC/CV target voltate, you could do this all manually now.
So we’re really down to software and connections to do the whole thing automagically.
Better, that same controller can swing both ways. You see there is no magic in rectification. A 3 phase 160 amp rectifier can be had on eBay for $45. The controller already has input capacitors. If we add a small box about the size of your hand to the car, we can also accept 240vac level II and provide that as an alternate input to this controller.
It automatically becomes both a Level II and Level III charger.
If a roundish video producer might suggest such a thing to a controller developer, say EVnetics, they could develop a very popular product from this, though probably of some expense. They sell a $2895 controller right now that would do it all with a bit of software. You might be tempted to add a big capacitor to the output, but it would frankly be a waste of time. The cells don’t care.
I would say the sticky part is if the EVSE really insists on 400v, and your components restrict you to 360v, those kinds of issues.
I have no inside information on what they DID with this input – only that I provided it. They may already “know” that it won’t work. If so, it is not beyond me to demonstrate that it does. I can both measure voltage and simulate a 0-5v output throttle signal with an Arduino. The software is almost trivial.
With the 300 ampere limitation, I would say a Soliton JR would be well within its means if there was someway to protect from an input overvoltage.
To actually USE a Level III EVSE, it becomes a bit more complicated. You have to communicate with the EVSE via a serial or CAN bus interface to signal it to turn on and off and so forth. So it gets to be a bit more complicated than I quite describe.
Can it be done? Oh yes. It’ snot really a thing at all.
Using the controller as a charger is a brilliant concept; I’ve not encountered the idea before. Like all the best ideas, obvious once it’s pointed out.
Re drag, I’d suggest that there are four significant types. Drag proportional to:
* speed squared (mostly aero, some transmission),
* weight (mostly rolling resistance, some wheel bearing),
* power transmitted (mostly transmission losses)
* fixed (such as brake drag).
I’ve had problems with brake drag in my GT40. A nuisance with a big V8, but pretty serious for an EV. Can be isolated by running on axle stands I guess.
Very interesting. So getting things to communicate and work automagically provides a nice hurdle as well as the 400 volt problem and the lack of ready made hardware. The fact that it can or could be done is reason enough somebody out there in the EV ether should make an attempt at it at some stage. Although, I guess it’s inevitable somebody (competent I hope) does.
Don’t forget a contactor or three to release the motor.
Interfacing a controller with a H bridge regen.. The software will be using your throttle in reverse while counting the AH.
It does sound simple but there’s so much money to be made out of the punters selling separate, disparate units all over the place. Especially unneeded electronics.
Actually I’m not advocating using your controller as a charger. But starting with a controller, you could fairly easily build a separate charger. They have a lot in common. It would take a LOT of contractors to disconnect the output from the motor, and connect it to the batteries, and disconnect the INPUT from the batteries and connect it to the EVSE.
But you could make a device that looked a lot like a controller, and charged your cells. And you could have a SEPARATE device for the AC rectification for home charging.
I did because as far as I know it is what is done with most if not all EV car manufacturers.
For the first guy to flog such a thing for us to buy will be on a sure winner at the expense of any rivals.
Chademo fast chargers do all the hard work. They are designed to output a specific current as demanded by the vehicle. Epyon have (or should I say purport to have) something called the Luna module. Its supposed purpose is to enable non OEM vehicles to incorporate chademo DC fast charing.
Imagine the depth of my surprise and diasapointment however when they won’t talk about it let alone sell one. More imaginary products. Gotta love this game.
I have read the chademo standard and canbus com message format documents. Not really difficult to implement. Someone will do it and make something like the Luna available to converters. Tempted to do it myself as we are supposed to be getting 30+ fast chargers installed in Ireland.
About drag.. I’m with Andyj on this one, and would also add the front wheel alignment in the equation. If the tire alignment is off, it can have a drastic effect on your rolling resistance. Doesn’t cost much to check it, if you haven’t already.
Hehe. I’ve just seen the next blog updates video.
It’s going to be conjecture time!
“190lbs of Noto meat” on a soap box derby made me laugh.
There are rumors some folks have used the regen function of a controller as a charger. I assume you’d feed rectified AC to the motor terminals, and it would feed the batteries like regen. You’d likely need to do some reprogramming of functions.
On the 0.2 Cd thing: Given that the most aerodynamic car ever mass produced, the GM EV-1, was 0.19, I highly doubt the Spyder is 0.2. Given < 0.3 is very good, I doubt even more. No need for doubt, appeals to authority, or whatever, however. It can be measured. Measure the front area (height x width will get you close enough), and measure the weight. Record speed vs. time several times each direction for coast downs. You can then calculate the Cd. Post up the numbers if you want help with the calculations.
I think ACPropulsion and the early Tesla’s use the controller and motor as a charger.
Had an idea from something I saw recently, you could use a torque wrench on a wheel lug to see how much force is required to move a vehicle manually and compare it to another vehicle. You could also use a torque wrench, an inch pound unit, on a wheel lug while jacked up to compare drag between each wheel, and compare to another vehicle.
I’m betting on wheel alignment issues on the eSpyder too. In second place would be something binding when the car has all its weight on its wheels.
How about the batteries of the Spyder?
How about determining if the drag is the same for the wheels. Do the soapbox derby test again without the car. Yes, just the wheel. Keeping it straight might be a problem but in an hour you will see what the rolling resistance is by how far one wheel type makes it down the street.
Also, a Push/Pull gage to move the car on a flat and level surface will also give you numbers to contemplate. Brian, did you ever freeze a part? Things drop into place when colder (smaller) No hammering is necessary.
As always – Great Show.
I may be completely off the wall, but could the difference be accounted for by the difference between the 3.44 and 3.88 rear end gearing?
There is about a 12% difference there, so the gears in the Spyder would be spinning in the transmission lube about 12% faster than in the Speedster.
How much that would affect the drag, I don’t know, but to me, the soapbox derby seemed telling, because only the transmission and rear end were involved, assuming front and rear alignment were good.
Maybe a thinner transmission lube would give some indication.
Anyway, terrific show as usual!
We have two possibilities going at the moment, and the most likely is this 3.44 vs 3.88 rear end.
As to transmission lube, we replaced Spyders with 75/90 synthetic, the exact same stuff we use in Speedster.
I have a suggestion.
Use a blank white paper and tape that inside the wheel arces. Where your most likely to rub against the tires. Then jump in both of you and go for a short run.
If your paper is black somewhere, your rubbing a tire.
It seems like you do, since you got a shorter run with the 180lbs of extra meat in the car.
But what do I know I cannot see from the film just how little clearance there is between the tire and the arc.
Just a suggestion of a very simple method to do the test.
Fast charging station – is it more than a CC source?
If it only has to charge to the 80% mark, then just charge at the set point supplied by the PWM from the car and stop when the car reach its maximum voltage.
It actually makes sense, to me, that the rear end gears are the cause for the mismatch. Even with a transmission in neutral the rear axle still spins the center shaft inside the transmission.
With the higher gearing in the rear of the Spyder, the center shaft is being spun faster than the one in the Speedster. Because of this both the center shaft and rear end receive more fluid friction from the lubricant. With the advent of the new tires having a smaller circumference, it makes the gearing spin even faster. This is evident from your runs with the new tires and increased RPM’s.
Doing the comparison is great. Making sure the cars are evaluated on as many of the same points as possible gets you closer to the real problem.
If you go far enough the Spyder will soon become a Speedster.
I have suggestions.
More Sun cream Jack.
*Jack up the rear, (suspension or raise gearbox) so the back tyres sit vertical and shafts horizontal.
*Fit wheels closer inboard so the load is nearer to the bearings to lessen any load drag.
*Granny suck eggs – Ensure no toe in or out front and back. (Toe-out is used to help self centre the steering).
*Finally, swap transmission if that 13% more rpm on the driven gearshaft actually shows more drag.
Out of interest have you not tried the soap box runs in top gear? This will give prior information on further losses down the line. More fun than noting stiction & rolling torques while on the workshop floor.
Suggestion; disconnect the drive shaft and do the soap box derby again. If situation stays the same, it’s the wheel alignment (or the rear differential itself), If Spyder improves, it’s the transmission.
Actually, I don’t know if that is possible with Porsche style configuration… I guess there’s no drive shaft between the tranny and differential, so have to disconnect whatever turns the wheels, if that’s possible with wheel support intact. Well, if I’m talking out of my behind, please ignore me.
The wheels are attached to the shafts.
One idea, likely already checked, occurred to me. Check the oil level in both gear boxes. The level can effect drag, it tends to drop as the level drops in most gear boxes. Of course, to low causes serious problems. Checking is free, and easy with a lift, so even if unlikely it seems worth ruling out.
Do all 4 wheels on the Spyder point in the same direction?
Are the Sypder and the Speedster built on same chassis? Beetle?
I vote for a four wheel alignment to start.
Also, the wheel bearing grease looked kind of stiff. Try some low temp moly from one of the majors, or maybe Alisyn MPG, which remains serviceable at temperatures as low as -80 degrees F to 400 degrees F and was originally developed for military aircraft high-speed turbine engine bearings.
Another suggestion to check wheel alignment.
Around 1:47:00 into the new episode it looks like the right rear wheel is not perpendicular to the ground but tilted inwards.
Nikita: I’ve already mentioned about jacking the rear for less inner side wall deflection.
Someone has already mentioned cells.. Maybe its the lack of a BMS while doing a soap box car run?
I also note for a fact if your tyres have white side walls they will be less “eco” than any silicon rich tyre on the market. So it stands to fact. White paint wrecks your mileage. Tried painting the Spyder Ivory?
Read about a junkie who used to smoke something. He said he feels like he’s doing 500mph sitting still. Maybe its the brand of smoke? This is where a BMS comes in.
Like most mysteries you start from one end and work your way through till you find it ain’t no other way. Enjoy the puzzle its the universe keeping us humble LOL Leigh
It does indeed keep us humble.
The mechanicals are much harder for me than the electrical. No really good meters on anything.
Prize for absolutely the best theory on this one goes to John Hardy of the UK. John notes that most of these cars originally had drum rear brakes and disk front brakes. (true). He further notes that on cars where the master cylinder was BELOW the calipers, the fluid would drain back into the master cylinder and you had to pump the brakes. The master cylinder on our Spyder is BELOW the calipers where on the Speedster it is not.
To cure this, they would often install a device called a residual pressure valve. This would maintain a bit of hydraulic pressure to the caliper or wheel cylinder. Wheel cylinders on drum brakes typically held more fluid and so they used a 10 psi valve for drums and a 2 psi valve for disk brakes. Ergo, if we had been upgraded to disk brakes on the rear but retained the 10 psi residual valve, the brakes would work fine, but have quite a bit of drag from the excess pressure.
I love this theory. The residual valves don’t work up on jack stands. They only activate after you bleed out the system and apply the brakes the first time.
Unfortunately, we have no sign of any residual valve anywhere on the car, front or rear.
But it was a GREAT theory.
Eric Kriss had a conversation with a transmission expert from the racing industry, who indicated that the R&P plays a BIG role in rolling resistance and lower (3.44) is better. This all fits.
We have an alignment scheduled for this morning. But I don’t’ expect much from that.
Lacking dramatic improvement from alignment, we are probably going to declare the R&P the culprit and just live with it.
Nobody has monopoly on ideas. Someone else will know what you already know.ok?
And nobody mentioned about BMS in this recent blog, only you. Is this you kissing Jack’s ass again? Just kiss it yourself but not at others expense.
Hello, you are the one who mentioned batteries.
We watched you kiss your own arse. xx
Depending on which VW Beetle master cylinder you are using you may have built in 10 psi residual valves, front and/or rear circuit. Checking should be as easy as putting the running car back on a lift and spinning the tires.
I’ve been perfectly happy with my old Beetle drum brakes and wide 5 bolt pattern, a vintage look. Still, those wheels and brakes you put on the Spyder sure do look cool.
Jacks already said they, (the wheels) spin freely.