Electric Motorcycle Armchair Engineers Discussion Thread. Truths, Half-truths, and Myths.

Discussion in 'Electric Motorcycles' started by T.S.Zarathustra, Mar 22, 2019.

  1. T.S.Zarathustra

    T.S.Zarathustra Been here awhile

    Joined:
    Apr 26, 2018
    Oddometer:
    365
    Location:
    Noord Holland
    A new thread for general electric engineering discussions about Electric Motorcycle (EM) truths and half-truths in one place where they are easier to refer to.

    To start the discussion. Here are few popular myths. Over the next few months, if I can find the time or energy, I will go through them and explain why they are wrong (maybe I'll add some more).
    1. If the manufacturer says it takes an hour to charge from 0-100% it will always take an hour to charge the motorcycle. #2
    2. EM are expensive because the batteries are expensive. #6
    3. To increase power all you have to do is increasing the current limit in the controller. #14
    4. Electric motors heat up because there is so much current going through them. #41
    5. Current flowing through motor will create resistance. #49
    6. Electric vehicles have the larger battery, but only the highest model has it unlocked to give the additional range. #56
    7. DC is more efficient than AC. #57 (#59)
    8. When battery packs are worn out they can be used in powerwalls. #63
    9. Electric vehicles have zero pollution. #133
    10. We are running out of "DynoJuice" so we have to switch to electric power. #200
    11. If you put a small windmill, or dynamo, on your electric motorcycle you can go further on a charge. #205
    12. Electric motors have full power from 0 RPM. #207
    13. Current equals torque, regardless of condition.
    14. Fast charge does not damage batteries.
    15. Charging every night does not damage the batteries.
    16. Torque in the motor comes from the batteries.
    17. Gearboxes have no benefits on EM, and are only added weight.
    18. Only current capacities of batteries matter, the low voltage can be compensated for by serial connecting more cells.
    19. Electric motors have full power on the whole range from 0 to 20000 RPM.
    20. To increase power all you have to do is increase the voltage.
    21. The power in EM is not in the motor, it is in the motor control unit.
    22. The power in EM is not in the motor, it is in the batteries.
    23. The only reason for bigger electric motors is so they heat slower.
    24. To get more power out of motor all you have to do is increase its cooling.
    25. If motor has built in fan then higher motor speed equals better cooling, and therefore cooler running motor.
    26. Power in electric motors is measured in Torque and is directly dependent on current.
    27. EM manufacturers will not match components of similar capabilities together, but will gladly purchase much better quality (and more expensive) types of certain components than is needed, so the customers can cheaply "tune" and get more power.
    Feel free to contribute. Please try not to use singularly unloquacious and diminutive linguistic expression to satisfactorily accomplish the egotistical contemporary necessity. If you think someone is wrong, please explain what you think is wrong. :thumbup
    #1
    Cybsal, Chaostrophy and ctromley like this.
  2. T.S.Zarathustra

    T.S.Zarathustra Been here awhile

    Joined:
    Apr 26, 2018
    Oddometer:
    365
    Location:
    Noord Holland
    1. If the manufacturer says it takes an hour to charge from 0-100% it will always take an hour to charge the motorcycle.

    In real life you will practically never have to charge from 0%. In most cases you'll be charging from 1/3 to 1/4 charge. Charging slows down between 80-90% and up to 100%. So a battery that will take an hour to charge from 0-100% will only take about half an hour to charge from 1/4 (or 25%) to 90%.
    #2
    mick, Cybsal, Chaostrophy and 2 others like this.
  3. ctromley

    ctromley Long timer

    Joined:
    Oct 18, 2012
    Oddometer:
    1,059
    Location:
    PA
    Agree completely.

    I'll add a little pertaining to why it's true for the benefit of the electrically challenged (because I can relate):

    If a lithium cell (for example) has a resting voltage of 3.7 V and you want to charge it, you need to connect a voltage source at a higher voltage to get current to flow into the cell. Typically battery chargers have constant current (CC) and constant voltage (CV) modes. Generally you start in CC mode, where the charger applies whatever voltage is necessary to reach its rated (or programmed) CC current level. As the battery accepts more charge, it takes more voltage to drive the same current, so the charger adapts and charging voltage rises.

    Any battery chemistry has a "do not exceed" voltage, above which bad things happen. (4.1 - 4.2 V for lithium ion.) When charging voltage reaches that limit (which generally falls around the above-mentioned 80 - 90% full region, depending), the charger switches to CV. From then on the current tapers off as charging continues, down to a pretty low value, which slows the process. A lot. See the graph at the link below for a visual. Note: I think (not sure) the CV time being roughly 3X the CC time is not typical of vehicles because the manufacturers don't allow fully charging or discharging a pack. This is to extend pack life. If anyone knows for sure, please confirm or correct me.

    https://batteryuniversity.com/index.php/learn/article/charging_lithium_ion_batteries
    #3
    mick and Chaostrophy like this.
  4. Madrodo

    Madrodo Adventurer

    Joined:
    Mar 18, 2019
    Oddometer:
    59
    Location:
    Belgium
    U / I = R

    U * I = W

    W * t = Wh
    #4
    CROSSBOLT likes this.
  5. Chaostrophy

    Chaostrophy Been here awhile

    Joined:
    Oct 24, 2017
    Oddometer:
    548
    Location:
    US
    I want quick charge and a dustbin faring (not that I care, but it needs much better than normal aerodynamics), and space for 4 bags of groceries. 300 miles at 75 mph as a hard maximum would be great (right now I'd want 230 real world, so 300 theoretical should be fine). Craig Vetter's 100mpg challenge bike would be a good starting point. Touring, not adv, sadly.
    #5
    RockyMtnRoadRash likes this.
  6. T.S.Zarathustra

    T.S.Zarathustra Been here awhile

    Joined:
    Apr 26, 2018
    Oddometer:
    365
    Location:
    Noord Holland
    2. EM are expensive because the batteries are expensive.

    Well, that was mostly correct 10 years ago. Today (IMHO) EM's are expensive because:
    A. They are mostly hand built. $$
    B. They are marketed as environmentally friendly. $$$
    C. They are made in small volumes. $
    D. They are basically luxury items.
    E. Customers are willing to pay.

    Batteries. In volumes of +1000, they have dropped in price about 90% per MWh in the last 10 years. Ok, manufacturers have added more batteries to increase range, but at current cell prices even doubling the amount of range will not greatly affect the price of manufacturing an EM.
    Speed controllers. There is similar situation about speed controllers. In a loosely translated Moore's Law, electronic components and microprocessors double in power, while halving in price, every two years. So in the last 10 years microprocessor and transistors, the main units in the electric system, have gotten 10 times as powerful at 1/10th of the cost.
    Motors are basically Teslas 130 year old invention so there is not much cost there.
    The rest of the motorcycle is nearly identical to Internal Combustion Engine "ICE" motorcycles so again, not much price difference.
    #6
    Chaostrophy likes this.
  7. ctromley

    ctromley Long timer

    Joined:
    Oct 18, 2012
    Oddometer:
    1,059
    Location:
    PA
    I think it's simpler than that. They're expensive because demand is low, for both good reasons and ill-informed thinking.

    A. They are hand built because sales volumes are low. Because of low demand. Gearing up for big production numbers to get the price down before demand has been established is a quick way to flush $Millions down the toilet.
    B. I'm trying really hard to see a connection between marketing something as environmentally friendly (and frankly, you don't see much of that from EM manufacturers), and how that somehow leads to a high price. Not seeing it.
    C. See A above, and demand.
    D. There's a lot of variation there, enough so I would hesitate to make such a broad generalization. They can be highly utilitarian for commuters. But yes, since motorcycles in general are luxury items for most people, EMs today are probably more so. But where they work, nothing works better.
    E. Some customers are willing to pay, but most aren't. The most frequent complaint about EMs on ADV, along with range, is the cost. Trust me, EM manufacturers are working Very Hard to reduce costs. Witness the Strike. Big volume is where the big profits are.

    It also seems to me that Moore's law only really applies when there is a strong demand driving development, so I'm not sure it applies to controllers. It works for phones and computers because people can't get enough of them and they're willing to pay for the hottest new thing. Today the primary demand for EV controllers is on the automotive side, mostly for development, not any great sales volume. (I believe Zero uses off-the-shelf or mildly customized industrial controllers, don't know about Lightning.) Are there other industries driving demand for FETs and IGBTs with killer amp capacity that are perfectly matched? For new controller architectures that can package more amp capacity in ever-smaller sizes? (EV Controller design is a seriously dark art at the currents they run - book knowledge doesn't get you even close.) I think Tesla might be the only one building their own automotive-specific AC controllers in high quantities. (That's a guess - anyone know for sure?)

    I do think Moore's law applies for batteries because there are so many applications that keep demand hot for them. Prices have dropped and performance has improved dramatically from years ago, but my gut tells me we're not quite at a break-even point with ICE and its related hardware. Anyone who has looked at cell prices to build a pack knows how pricey they still are. I could easily be wrong though (hard to know what ICE costs are to the factory). It would be interesting to see an analysis. Of course, Moore's law dictates that even if I'm right today, it won't be long before I'm wrong. I believe electric power will beat ICE on cost. Can't say when, but I'll bet it happens a lot sooner than many here think. There is a lot of progress yet ahead of us, and a lot of cash and talent chasing it. The level of advancement of today's EV powertrains is really little better than a cordless drill.

    The low demand / high price conundrum is a problem in the early days of any emerging technology. It's one of the main reasons subsidies are needed for EVs, renewable energy, etc. to get them off the ground when established, mature competition is already in place. Computers and mobile phones didn't need them because they both created their own brand new markets - there was no real competition for either when they launched.
    #7
  8. PeterW

    PeterW Long timer

    Joined:
    Nov 11, 2005
    Oddometer:
    4,373
    Location:
    Gold Coast
    Computers and early phones were insane expensive. Even more differential than the EV's have now.

    The early IBM PC's cost more in the dollars of the day than they do now, not sure of the exact translation but likely 3x more in todays dollars than you'd pay now.
    "
    1981. The PC 5150 was IBM's most successful attempt at a personal computer at the time and was used as the basis for most computers that followed. The basic unit sold for $1,565, and the full model for $3,000.
    "
    An online inflation calculator suggest >$8,000 in todays $'s and a PC now is <$1000.

    Hopefully the price will drop seriously as standardization kicks in. Battery cost aside they are MUCH simpler than an IC engine and should end up far cheaper once they become dominant. I'd be hoping for an around 10x price drop with economies of scale and simpler engineering factored in. (Adding back greed ;), maybe only a 3x drop).
    #8
    Chaostrophy and ctromley like this.
  9. liberpolly

    liberpolly Nu, shoyn, nudnick!

    Joined:
    Sep 21, 2015
    Oddometer:
    5,211
    Location:
    Seattle
    The best remedy against greed is competition. Once demand grows and volumes catch up, EM will be on par with the price of the equivalent ICE.

    But I'll still miss the sound :/
    #9
  10. Bt10

    Bt10 Long timer

    Joined:
    Dec 18, 2012
    Oddometer:
    1,057
    Location:
    Saranac,MI
    Any info on the axial induced motors?
    #10
  11. ctromley

    ctromley Long timer

    Joined:
    Oct 18, 2012
    Oddometer:
    1,059
    Location:
    PA
    The only common motor I know of with "axial" in its name is axial gap PM motors. If that's what you mean, they're available from several sources. Their main advantage is less weight for the power produced, but cooling can be a challenge. And they can be pricey. In the DIY renewable energy field, there are lots of vids and other instructions for making your own axial gap generator for wind turbines.
    #11
    Chaostrophy likes this.
  12. T.S.Zarathustra

    T.S.Zarathustra Been here awhile

    Joined:
    Apr 26, 2018
    Oddometer:
    365
    Location:
    Noord Holland
    Sadly, axial induced motors are not as good design as some would have you believe. The problem lies in getting even magnetic field in a piece of metal that is basically shaped like a slice of Pizza, or trapezoid. The pointy bits are partially surrounded by wire, and will be oversaturated (causing heat / power loss), while the bits in the center are not close to any wire, so they will be undersaturated (and are therefore unneeded weight).
    #12
    Chaostrophy likes this.
  13. Bt10

    Bt10 Long timer

    Joined:
    Dec 18, 2012
    Oddometer:
    1,057
    Location:
    Saranac,MI
  14. T.S.Zarathustra

    T.S.Zarathustra Been here awhile

    Joined:
    Apr 26, 2018
    Oddometer:
    365
    Location:
    Noord Holland
    "3. To increase power all you have to do is increasing the current limit in the controller."

    This one is partially, maybe 10% true. You might get away with 5-15% power output from the motor by increasing current limit. But most systems are designed with each component in balance with the others, so you do that you are likely to be dangerously close to the limits with at least one of the other components. Back when I was designing electric drivetrains I tended to slightly overspec the motor controllers, and then limit the current and motor speed in the software. This was done primarily to increase reliability in the most sensitive, and expensive piece of electronics, and allow it to run cooler. My ramp up and down times (soft start and soft stop) were also set as high as possible (within reason), to prevent damage to gears, movable parts in the machines, and avoid excessive yoyo-ing in machine speeds. Increasing the controller specs from 100 to 120 Amps is a insignificant price increase and is usually offset by better reliability. But I didn't overspec it so much that you can just increase the current limit no end, and it will run fine if you just slap a fan on it.
    The 90% not true, in this claim, is that while you can increase the limits in the controller, it is unlikely that the motor will be able to handle much more power, or the batteries be able provide much more current, without damage. Even if you slap a fan on it (provide active cooling). Using fans is something which most experienced designers try to avoid at all costs, because fans are easily damaged and have a limited lifespan.
    #14
    Chaostrophy likes this.
  15. ctromley

    ctromley Long timer

    Joined:
    Oct 18, 2012
    Oddometer:
    1,059
    Location:
    PA
    I'm going to disagree with you strongly, and we'll both be right. My previous claims that increasing current limit will provide a commensurate increase in power are based on the assumption that the other components in the drivetrain are up to it.

    It's also worth reminding people that my claim has been made in the context that it's better to increase current on an EM than add a transmission. A transmission is a torque multiplier. But on an EM you also have the option of multiplying torque electrically. (And bumping up current limit isn't even the only option for electric torque multiplication.) As it works out, for the magnitude of bump needed to compensate for no trans, and especially when working with a clean-sheet new design, the other related effects and opportunities that present themselves by increasing current limit generally make it a better choice than adding a trans. (Making space for more batteries is one of the major opportunities.)

    But back to your more general statement. To my knowledge, the torque equations for every type of electric motor show torque being directly dependent on current. (I know for sure this applies to series wound, PM and AC induction. It gets more interesting for separately excited, compound or some hybrids.) There are other factors in the equations, but they are constants for any particular motor. (For more 'interesting' types, the additional influences can be variable to optimize performance.) You must address the torque equations when you get to #5 on your list. Double the current and you WILL get double the torque, assuming of course that you've done your homework and ensured that the motor doesn't turn into a ball of copper snot or self-destruct in some other way. That is generally a question of cooling and time. And the battery pack needs to be OK with all the current the motor can draw and the controller will pass. (Typically not an issue with long-range packs.)

    There are guys today running series motors at the drags running 10X their rated current. Ratings are generally based on continuous use. Drag cars run for around 10 seconds. Series motors aren't light, so they'll soak up a lot of thermal load before they get hot. The weak points are typically brushes and commutator bars. Make sure those are up to it, and 2X the current WILL give you 2X the power. It's simple math. The motor equation says torque is directly proportional to current. Power is directly proportional to torque and rpm. 2X current = 2X power.

    Disclaimer: While fundamentally true, this ignores some lesser influences. The series wound motors involved can handle that kind of current briefly, with some basic but critical mods. Longer with more cooling. But bigger currents and over-volting (generally no more than 2X) can warp the magnetic fields. This can be mitigated by adding compensating coils and/or re-timing the commutation, but you won't get quite the full 2X increase for 2X current. Close, but not quite. PM motors are more limited. Go too high on current and you can de-magnetize the magnets. Depends on the particular motor you're leaning on. (And honestly, I have no experience here, so I don't know how one hot-rods a PM motor.) But the basic relationship is that torque depends directly on current. Most motors have, or can be given, some headroom for more current.

    I would hope that it's obvious to anyone that an increase in power from one change needs to be accommodated by all other system components subjected to the increase. Battery packs can theoretically be configured to provide any current and voltage you want, though size and weight can be limitations. (Generally speaking, if you have enough kWh for comfortable range you have plenty more current capability than needed to feed your motor. But the BMS might need a code tweak to allow more current.) Drivetrain loads are primarily limited by traction, regardless of torque. (The tire contact patch is a wonderfully effective mechanical "fuse".) Torque-induced shock loads (and trust me, electric torque can produce shock loads the ICE guys don't see) can be reduced with damping and/or software (i.e. current ramps). Motors can take more current to an extent, but cooling might need improvement. Elsewhere, smdub explained how the 75 kW (continuous) EV2000 drivetrain works with a motor the size of a coffee can - by going to extremes with cooling:

    https://advrider.com/f/threads/lightning-strike-s-twice.1362769/page-3#post-37268928

    I think where our views primarily diverge is that you're coming from a motion control background and I'm coming from general transportation. (It can also be seen as kind of a hot-rodder vs. engineer dichotomy. I've been both.) You work to a System Requirements Document. Your practice of using a 5 - 15% factor of safety in a well-defined environment makes lots of sense. The system requirements for general transportation are so severe, diverse and even unpredictable that any particular user is using a very over-built system. When Leaf drivers in Arizona and New Mexico saw rapid aging in their packs due to hot weather, people in Canada could not have cared less. But when Nissan fixed the problem in production, Leafs in Canada got the fix too. Canadian Leafs are over-built. And Arizona - New Mexico Leafs were always over-built in terms of cold performance. (Problems like that are supposed to be caught in testing, but sometimes they slip through.)

    The example that strikes hardest at your view, and strongly supports mine, is the Zero SR. A few years ago, SRs were going into thermal cutback when ridden hard at high speed. I believe the SR is identical to the S in terms of motor and battery, they got the SR's additional power by bumping up current limit. (Zero experts please correct me if I'm wrong.) Zero is a small company and apparently didn't test well enough before releasing the SR. On later models the problem was fixed, presumably with improved cooling, because the major components remain basically unchanged. Which is exactly what I've been claiming. Zero just got caught by not verifying the other sub-systems were up to snuff when they increased current limit.
    #15
  16. smdub

    smdub Adventurer

    Joined:
    Oct 5, 2017
    Oddometer:
    84
    Location:
    Maryland
    If a motor is designed right, you can not get 2x increase in torque when you 2x the current. The steel should be running near saturation in an optimum design. Can only generate so much flux across the airgap given some number of Amp-turns. It quickly goes past diminishing returns. If you CAN get 2x the torque for an increase in current than it can be argued the orig motor is WAY too big.

    The series DC motor dragster guys don't care too much about the efficient use of power;)

    You are correct that too much current in a PMSM also risks demagnetizing the magnets. Also, Technically you can field weaken a PMSM (run current tht generates a flux to oppose the magnets. Lowers the back-emf and you can spin it faster.) We did this on the V22 Osprey APU starter. *BUT* it comes w/ a large risk. If something goes wrong and the controller shuts down, the magnetic field instantly goes back to full strength and the back-emf jumps back up. Quite likely more than your DC link can handle. An inverter works as a 3-phase rectifier in reverse. The inverter goes POP and tries to stall the motor! Can't remember if they ever snapped a shaft testing when that happened. We had to put a contactor between the inverter and motor to protect against this.

    Careful, I said the EV2000 *ROTOR* was the size of a coffee can. The wound stator is around that and added a couple inches. The whole thing was much smaller than a beer keg. Mostly all aluminum housing w/ fins and cooling passages. It was so much smaller than an ICE there wasn't a huge effort to minimize the shell. Wonder if I still have a pic of it somewhere...
    #16
    magnussonh, Chaostrophy and kiwial like this.
  17. ctromley

    ctromley Long timer

    Joined:
    Oct 18, 2012
    Oddometer:
    1,059
    Location:
    PA
    Those guys aren't big on theory. They try stuff and see what works. They're very good at invention, and test runs only take 10 seconds. (Sometimes less!)

    They use series motors because good ETs are very dependent on a killer launch. That takes a lot of torque. If you look at the series motor curve you see that as you approach zero rpm, torque goes to infinity(!). In practical terms, since current is proportional to torque, if you have a controller that can deliver monster amps, you get a monster launch. I read about a guy that had a rail with Top Fuel driveline parts, but they'd keep snapping when he dropped the hammer. You find limits, you push past them.

    Like the fact that series motors can't spin too fast thanks to the commutator, and they tend to run out of steam at high (for them) rpm. So they figured out that you could use two smaller motors, run them in series for the launch (same current runs through both motors - two for the price of one!), then when they spin up and can no longer draw big amps, use contactors to quickly switch them to parallel (more volts for each for bigger top end). It's effectively a fully electric transmission. Anything to get that launch, then whatever else is needed to make it work elsewhere.

    There's one guy who has a Datsun 1200 sedan with about 1100 hp worth of lithium in the trunk (which also happens to be good for maybe 120 miles of range) and a clever custom motor with two commutators. It's essentially two complete motors built on a common armature shaft and weighs probably 130 lbs. The car runs DOT street tires on 13" rims in non-tubbed wheel wells, has a nice interior and operational doors. Street legal, licensed and insured, does great grocery runs, docile as a lamb. He drives it to the local strip and runs the 1/4 mile 10.2 seconds. 0 - 60 mph times are around 1.8 seconds.

    Efficiency? The hot-rodder will tell you that concern over efficiency is misplaced. It's something you'd expect from an engineer. 'Does it work better than anything else?' THAT's the pertinent question. And in his world, he's right.

    It is absolutely essential that we understand theory to advance. But we can't let it be a limitation. If you do that, invention dies and we miss out on some great stuff. Sometimes breaking the rules is the best thing to do. You need to understand when theory might be an incomplete description of reality. How the end result works is what matters.

    In theory, theory and reality are the same. But in reality, ....

    EDIT: BTW, that 1100 hp battery pack has a capacity of 22.7 kWh. Given his range of 110 - 120 miles, that calculates out to an average of around 200 Wh/mile. That's pretty damn good efficiency when driven in grocery-getter mode, more than comparable to any electric grocery-getter you can find.
    #17
    Chaostrophy, MJSfoto1956 and jimmy650 like this.
  18. T.S.Zarathustra

    T.S.Zarathustra Been here awhile

    Joined:
    Apr 26, 2018
    Oddometer:
    365
    Location:
    Noord Holland
    For various reasons I believe this article to be a hoax. It looks like it's written by someone who is familiar with buzzwords and electric formulas, but does not understand the functions behind it. I am not saying that the idea is hopeless, just that I find it unlikely to work as advertised. :-)
    It is carefully worded with big claims followed by buzzword explanations.
    There are only renders of the motor, not pictures.
    They say "permanent magnets are farther away from the axis, resulting in greater efficiency and leverage around the central axis.". But they are in effect moving the Magnetic Field (that does the work) from the edge, and closer to the centerline, which will reduce efficiency. More mass further from the centerline will also make the motors slower to spin up and slow down, resulting in less efficiency (if you take a stick and put some weights in the center it will be easy to spin end to end, if you put the same weights near the end it will be much harder to spin).
    They say "Magnax recommends 750 V from the battery". At 750 Volts, electricity will jump over 1/12" spark gap in best conditions. It will jump a much larger gap if the conditions are bad, like in damp or ionized air. The power connectors in the drawings look awfully close to each other for that voltage, not mentioning that the insulation in most wire for electric motors is only rated for 500-600 Volts.

    Have a look at this article from ABB, a long established electric motor manufacturer, where they get over 99% efficiency from standard design electric motor. https://www.abb-conversations.com/2017/07/abb-motor-sets-world-record-in-energy-efficiency/
    #18
  19. Bt10

    Bt10 Long timer

    Joined:
    Dec 18, 2012
    Oddometer:
    1,057
    Location:
    Saranac,MI
    Thanks for the article.
    #19
  20. ctromley

    ctromley Long timer

    Joined:
    Oct 18, 2012
    Oddometer:
    1,059
    Location:
    PA
    A hoax? Really? You seem to believe axial gap motors are some kind of vaporware scam. They're not, something that is easily verified by searching on "axial gap motor". This type of motor has been around for decades. It tends to be larger diameter, but shorter axially, than an equivalent radial gap motor. It spins up just fine, thankyouverymuch. In fact it is comparatively so light and compact you can easily fit two on an EM. The TTXGP (now TT Zero) race at Isle of Man has been won repeatedly by bikes running axial gap motors of the Lynch type, sold by Agni (now owned by Saietta Group, top motor on page 3):

    http://www.saiettagroup.com/downloads/SaiettaGroupIntro.pdf

    The reason there are only CAD renderings is that these motors are very application-specific. There are no stock models. If an OEM is interested, proof-of-concept prototypes (typically not brochure-quality stuff) are demonstrated. If there's interest, a short trial run is made to the client's specs for evaluation, and they may have only a passing resemblance to the generic CAD renderings.

    Your statement, "More mass further from the centerline will also make the motors slower to spin up and slow down, resulting in less efficiency" is misleading. I do understand that you're responding to claims in the Power Electronics article, but those articles are nothing more than press releases written by marketers. (That one was written by a co-founder of the company.) Marketing speak is known to be, and therefore should be expected to be sloppy. You, being an engineer, need to be more precise. Motor efficiency specs have nothing whatsoever to do with how quickly a motor changes speed, and any differences in rotational inertia between axial and radial motors are likely to be greatly overshadowed by that of whatever they drive.

    I would also hope you know that there are connectors and insulation rated for very high voltages. And that any expense they add might easily be made up elsewhere due to the lower currents that high voltages allow. It all depends on the particular application at hand.

    Efficiency was obviously a very high priority for the contract job that ABB built those record-setting motors for in your link. I'm just as impressed as you are at their efficiency. But efficiency is not the supreme measure for motor effectiveness. In most applications there is a very different and much more diverse mix of priorities, which makes for a dizzying variety of possible solutions. Thinking rigidly in terms of theoretical ideals is a tempting way to narrow down your choices, but it masks opportunities.

    And BTW, maximum efficiency has, by definition, relevance at a single point. Another advantage of PM axial flux motors is that they maintain their efficiency over a broad speed range, a very critical requirement for vehicles that may be less critical in the record-setting ABB motor. Real-world use is what matters.

    Understand the limitations of theory as applied to reality (which varies wildly from theory to theory, and also the application at hand), then treat engineering as the creative endeavor that it is. The art is in the compromises. That Datsun 1200 I described elsewhere is one of the most masterful, creative - and effective - collections of compromises I've ever seen. A set of compromises that theory-centric engineers would dismiss as unworkable on paper - so they wouldn't try.

    Another example: In the '90s I was working at a manufacturer of aircraft instrumentation that was supporting Cessna in their bid to win the JPATS contract, producing over 700(!) small planes to all the flying branches of the US military as jet trainers. It was a VERY lucrative contract, which is why seven different companies developed new planes in hopes of winning it, knowing that six of them would be wasting a lot of time and money. (Cessna didn't win, but they ended up with the sexiest little sport jet you could imagine. Didn't do anything with it, not enough market.) The winner was Beechcraft - with a prop plane! They so successfully mitigated its 'proppiness', and simulated jet behavior, that it flew exactly like a jet. It fulfilled its main purpose, with the added benefit of very un-jet-like purchase and maintenance costs.

    That was a very creative way to reach a surprisingly effective compromise, based on an apparent non-contender, against some very capable competition.

    Theory is a useful - and essential - tool. But it's not enough. You need to understand what theory can - and cannot - tell you, and why. You're not running with the big dogs until you have a real sense of what works, what doesn't, why, and the limits of stretching or modifying capabilities - in the real world. Great solutions don't come from textbooks. They come from creative minds.
    #20
    Chaostrophy likes this.