Serial-hybrid Delta SWB Velomobile

xenodius

10 kW
Joined
Sep 3, 2012
Messages
551
Location
Florida
So over the next few years I hope to complete this project. This'll be my build thread. Should be fun and full of trials and tribulations, but I have a plan to streamline parts of the process and I have a good idea of how everything will fit together. Any input is definitely welcome, this idea has been going under mental revision for at least a year now.

I love my KMX/A123/Cro/Max-E build but I want something that's more comfortable, weatherproof, and has more storage. Then it'll be perfect...

My goals? Minimal maintenance, maximum comfort, maximum versatility, uncompromising performance. Too good to be true? I hope we find out! It'll have a Cro, Max-E, serial-hybrid drivetrain-- that is, regenning BB with a PAS, a seat that is either recumbent or pops up with the handlebars for motorcycle-like straddling that should even accommodate extra passengers, zero point-turn radius, awesome full-suspension all around, a geometric and possibly an additional pneumatic damping system to limit rapid tilting for safety in low-traction conditions. Oh yes, and I think I will get 800x Samsung SDI 25R cells, 2.5ah and "10c" rated, overblown number but they can manage 1c rapid charge and if the whole pack can discharge 80,000+ watts peak in ideal conditions, I think I can expect passable performance in the winter. With a 1-hour charge on a 200-300 mile pack, I should be able to ride across the country. Not to mention that I'll still be able to charge overnight off of 110v if I have to stop somewhere and they don't have J1772/NEMA 14-50.

I'll achieve these points with the following design elements:

Vertical leading link fork. This will facilitate several things-- firstly, its a good design for a suspended FWD hubmotor, very sturdy. Second, the geometry works well for zero-point turns, which will remain ergonomic via simple U-joint linkage. This linkage will also allow easy adjustment of handlebar height, which will facilitate yet another point; the back of my seat will be able to flip up near the headrest and lock up, so that I can ride upright if I want. At least part of the fairing will be attached to the handlebars, and raise with it. There will be doors, which will both move forward on the inside of the fairing

Short wheelbase. The front wheel will be vertically in line with my knees in a reclined position, this will increase load on the front and help with traction. This will also help in slippery conditions, and make those aforementioned zero-point turns quite negotiable. I plan on regen-braking a lot but I'm looking at getting a moto hydraulic brake and disc on my Cro... alternatively, some DH brakes should do.

Self-centering geometry. The rear swingarms will be designed such that leaning will increase the overall center of gravity as it moves from the midline, instead of lowering it like a traditional linear track vehicle. It probably won't be enough to reliably remain upright at a stop, I'll have a lean-lock for that, but it'll help dampen excessive leaning due to loss of rear traction (which should never occur anyway in a FWD setup-- if you can tell, I plan on riding in snow and ice) Furthermore, It'll be very easy to weld up spare, xtra-wide swingarms if I decide i need more track in the winter.

Full fairing with a bubble. I want to be able to hunker down and cut through the wind. I don't want to blink when I think about riding in -20F weather. Or 100F weather, I've already designed some NACA ducts that are basically drag-free, in fact, I'm thinking of printing a "Cromotor Wind Tunnel®" with a similar design, that will basically pump air through the motor. I'll need some good hardware, and I'll probably use a latch to hold it forward and some 50lb magnets to hold it against the seam strip when its pulled back over me.

Serial drivetrain. I plan on riding at speeds from 5mph among pedestrians up to 60mph on the highway, and if it's cold, I'd like to stay warm that whole time. I plan on using a reversed mid-drive and cheap 6fet for resistance and energy. It won't be as efficient as a chain drive, but I don't care about that with a 8.4kwhr pack-- I care about staying warm, and getting exercise, and this accomplishes that better than a complex chainline on a FWD leaner... Also, I really want it to be super quiet, and chains can be noisy. A single belt drive or maybe BAFANG should be pretty quiet though. And it should make PAS perform much better, as the resistance is constant regardless of your speed-- zero, or 60. So it'd operate almost like a torque throttle. It will also make for a less cluttered dash-- I'm going to have handy breakers for switching panels and pedals between the pack and the coils, and for many lights, speakers, and other accessories. And things add up fast. I'll probably have a switch to enable/disable PAS too.

Big ass trunk. I want lots of convenient storage, and a delta gives me room, so it'll be big. And you can bet that empty space under the flat bed of the trunk and behind the seat will be full of batteries. Since it'll lean, things won't even slide around. I'll have room in the front of the fairing too. I'm thinking about putting together a FOB for this thing, possibly Arduino based... if I do, then I would definitely

Ylh03oil.png


The fairing, and charging. Based on the highlighted green surface area of the first top I sketched out, I have 6400 square inches that I can put panels on. Round that down to 4000 for logistic/optimization/wasted space and I can fit eight 50W semi-flex monocrystalline panels. Now I don't expect 400W all the time but 200W alone would be absolutely crazy awesome, and a more realistic figure-- it won't all get direct light at once, anyway. I also might change the fairing design a bit to accomodate the panels, and I'll be running flow dynamics to minimize turbulent points. When I get some of the ergonomics of the frame under the beast in my model, I'll make a better fairing model with inlays for the panels; then I'll lay up carbon fiber over the whole printed mold as usual, but before I put any vacuum layers on I'll drop the panel on and route the cables through. I would think 400W of panels should be able to manage 300wh total even on mediocre day... and that's enough for my daily commute. And that one day when I get perfect sun and 400 watts out of them, and I'm pedaling up another 100+, and I'm averaging 30mi/h indefinitely... That's when my EV grin will hit my earlobes. =)

For wall charging, I think I'll build myself a 100 amp coil and get a few Eatons, and make a nice system behind a little flap in the fairing with a self-spooling 110V cable, NEMA 14-50, and J1772 routed right through the NEMA-14-50. And a big heatsink for my Max-E. With Samsung SDI-25R's I can rapid charge in an hour, should I need to, and the rest of the time they'll get a loving .25c charge and rest at a comfortable 50-70% charge. I should be able to go well over 300 miles if I economize and/or pedal, and should be able to ride 200+ miles with abandon (but probably not field weakening.)

I know it's a lot to pack into a single vehicle but I don't see why it can't be done, as I've seen examples of all these bits by themselves. I'm about to get an Ultimaker 2 (i think, bartering) and it's capable of producing exceptionally precise (~20 micron resolution) FDM prints when tuned properly, so I could print a nice dash in Woodfill, my fairing mold, aero side mirror mounts, headlight mounts, NACA ducts, fenders or molds for composite fenders, etc... even nylon bushings or bearings, for small hinges and so forth. With such high resolution, it's easy to glue ABS parts together really solidly and get a nice finished piece with no visible joints-- after a little acetone vapor, anyway. Plus with infill you can make it lighter than sheeting...

I want it to serve as an all-weather vehicle with worry-free range, silent and comfortable operation with excellent suspension, lots of cargo space, good performance and handling, and integrated solar charging... for my ultimate goal; a vehicle that I can operate for 3,000 miles without plugging in to a wall, changing a tire, checking a dipstick, anything other than verifying things are working as they should. It's an arbitrary goal and bragging point. Of course I'll have longer term maintenance items with bearings, U-joints, suspension, and other linkages... but weather protection will go a long way in preserving those. I would like to have the frame done within a year, and the fairing done the next year. We'll see. I've been really really motivated but grad school is ridiculously time-consuming and I need to get a welding table before anything else, I think... large steel plates are expensive. I'll probably end up spending at least 5k all together and I might cannibalize my previous build too! But, I want it to be my transportation endpoint. Extremely custom, comfortable, reliable, and versatile.

I want it to be better than a car.

More models and simulations will come in the future... notably, the frame, and a more realistic fairing-- more teardrop-like. Which shouldn't actually take too long to put together as I already have sketches drawn up for many pieces. I'll also be printing basic jigs to accelerate the framebuilding process... using SHS for most of it. It will NOT be monocoque-- aint nobody got time for that!
 
Damn! :shock: Why not. Go for it. Cant wait to see the progression. Obviously, if you need any help just let me know.

Cheers,
Brake
 
Yeah long post. I've been stewing for a while, flip-flopping between a bunch of different options... and I think I've settled. I really want a leaning delta. No carpal tunnel for me!

I really appreciate that and will take you up on it! Right now I'm still getting used to Solidworks, it's the most daunting bit I think, it's very powerful and what I think I need to use, but doing these complex curves with insets and limited angles and then aerodynamically optimizing it all will take a lot of hours. I've made about five fairing mockups already-- I know I should start by modeling the frame first, I just like to work on the parts that I'll get to see so I can get excited =) And I still have some decisions to make about exactly how I want the fairing shaped, especially with PV panels on there that have limited flexibility. I also want to have a large rear trunk, but still have an efficient tailbox... Design is full of compromises...

I wish I could find the video, but there's a European guy who put together his own electric velomobile and built his seat such that the back popped up, hinged at the headrest, and locked upright so he essentially had a motorcycle saddle. He could easily switch while riding-- I don't expect I will be able to so that easily, but I think it'd be a simple addition and if I ever ride on trails or perhaps around town in the summer heat, I may make use of it. With 300 mile range and 1 hour charges I may find myself on some expeditions...

You don't happen to have a pump for vacuum bagging, do you? That's an expensive tool I'd probably use only once. No time for all that homebrew like I'd hoped, but I'm more excited about this! The more I look at it the more I realize how much room I'll have for a battery pack and accessories and extra power supplies, e.g. Eatons. I will also need a compressor to blow a Lexan bubble, but I should probably get one now that i have room =) And I might bother you for a pipe bender for the frame of the seat. But I'll cross that when I get there. =) I think I'll also repeat the KMX approach of ABS Sheet + Cut Generic Foam + Cinching Mesh = SEAT. It's easy to clean AND it breathes well. If you need anything printed let me know, I'll be doing nylon/ABS/PLA/Woodfill to start. I actually thought about trying to print the seat, and I may still do that actually, with a simple frame underneath it or through it-- with the right infill these parts can be pretty darn light and pretty darn strong for what they are. I guess you can machine what you may need, but with some plaster, wax, and a forge we could cast anything we want out of metal with reasonable precision, if we needed to!

I'm probably going to print and paint or layup the smaller bits too, like streamlined rear-view mirrors. I plan on streamlining the swingarms and fenders with 3d-printed pieces, too. I can't wait to find out what the efficiency will be-- I'm not betting on an increase over my KMX, but I know based on comparisons that it should be about twice as efficient. That's just too good to be true, as that would mean 400 miles at 40mph...

Did you ever try that Parabola? I hope it wasn't too big :twisted:

EDIT: My most recent-- and favorite-- fairing mockup. Not perfect but I think it will give me the best combination of space, aero, and solar real estate. Oh yeah and it's only 1/4th of a fairing, LOL. Worth mentioning now, because of the taper it's longer than might be absolutely necessary. But I think the aero improvement is worth it, I'll have to actually check. Turns out it's way easier to work with as a solid, vs. a surface, it should be very easy to add headlights and so forth once I pick some out.

WzPhxP3.png


Jq4h3uH.png
 
Saving the Firestone for a special occasion. Like when I get home:)

I don't have a vacuum pump like that. Next tool is a spot welder. Already have the parts, just need time to build it.

What a coincidence, I'm going through the Solidworks learning process right now too. Found some cool stuff on grabcad already done. I'm used to designing it all I'm my head, i think Solidworks will be fun after getting used it. Looking forward to the FEA stuff that can be done afterword. Mid mounted hub is next for me. Full downhill style. Adappto and mxus 3k. Toying around with the idea of a mid mounted 3 speed hub for gearing. We will see.

.
 
It'll be good for years to come =)

I've been doing more reading and apparently some people make a mold, heat up their lexan blank in the oven clamped in a wood frame, then take it out and carefully lay it over their mold... most people use blow them in a bubble or vacuum over a mold, but they're not doing one-offs. I'll have to read about techniques a little more. But that'll be the very last thing I build-- first the frame, then the fairing, and so on...

Oooh, for battery tabs, or a bigger one for sheeting...? I was thinking of building my own battery tab welder, I'll have about 800 cells...

There's a lot of cool stuff out there. It's definitely different, like speaking a new language... it's already getting more fun! FEA is very cool, I probably won't be able to use it much here because carbon fiber is anisotropic, and my print/mold will have high infill (mostly hollow, honeycomb pattern) that isn't modeled in solidworks but in Cura when I slice it into layers before printing. I can't wait to really use the Flow package...

Ohhhhh, my friend, I am so jealous. I can't justify it yet, but a downhill bike with a super torquey mid-drive would be an absolute blast! Really requires balancing a lot of factors for performance. I plan on having lots of suspension travel but this bike will be a road machine first. I can't wait to see it, I'm sure it'll be beautiful! Excellent motor choice. It would be cool to extend the gearing range a bit. Curious what 3-speed hub you had in mind that would take the punishment...? I thought they were all chintzy over ~1kw but haven't looked in ages. I've been tempted by lightningrods mid-drive kit and even the Bafang middie, but a mid-mounted hub like that is definitely better for off-road, I think.
 
Did you get a 3d printer? There are so many different ones. It's amazing how far they have come and the prices are so reasonable now. I'm amazed at how you still have time for this stuff while in school.

The spot welder it for batteries. Hope to get it working after I get back. Capacitance discharge style. Soldering those batteries takes forever. It's all yours to borrow for those 800 cells once its up and running.

I have to figure out the gear ratios first to see if it is viable, but the hub will be used differently. The motor would spin the outside of the hub. The outside of the hub would spin the rear hub. The gearing is only for peddle input. If possible. The main thing is that the power from the motor doesn't go through the gearing. It won't be a high speed bike, low speed and torque. Plan is to keep regen braking too. ;)

Those are the goals. Still working it out. Maybe high speed schlumpf too. $$$

Would love to see some more pictures of your plans. The frame and suspension sound really cool.
 
As far as IGHs go, I think that as long as you have something to keep the slack taken up so there is no lash in the gearing, like something that just lightly pulls the chain tight (and thus the IGH's gears) just before applying the motor power, even the cheaper ones would probably be fine.

Based on the disassembly pics and whatnot of failed ones we've seen here on ES so far, the failures were of teeth being ripped out at the roots, which then get ground up in the rest of teh gearing to destroy other things. Mostly that seems to happen from lash, where the gaps between teeth mean that teeth are slammed against each other whenever power is suddenly applied, and the hammering will eventually break something.

If there's no hammering, but just smooth application of the torque, it shouldn't cause that kind of a problem.



Regarding your fairing, I like the most recent version: it reminds me of the Lotus 7 from Secret Agent / The Prisoner, which is one of my favorite of car designs. Oddly enough, someone on another forum recently suggested I make the new bigger doggie trailer like that. :)
 
Brake said:
Did you get a 3d printer? There are so many different ones. It's amazing how far they have come and the prices are so reasonable now. I'm amazed at how you still have time for this stuff while in school.

The spot welder it for batteries. Hope to get it working after I get back. Capacitance discharge style. Soldering those batteries takes forever. It's all yours to borrow for those 800 cells once its up and running.

I have to figure out the gear ratios first to see if it is viable, but the hub will be used differently. The motor would spin the outside of the hub. The outside of the hub would spin the rear hub. The gearing is only for peddle input. If possible. The main thing is that the power from the motor doesn't go through the gearing. It won't be a high speed bike, low speed and torque. Plan is to keep regen braking too. ;)

Those are the goals. Still working it out. Maybe high speed schlumpf too. $$$

Would love to see some more pictures of your plans. The frame and suspension sound really cool.

Yeah time will be an issue. But I'll have a little more of it in the summer. But I find this sort of thing cathartic. It's going to be my only hobby for a while. And Modafinil will not be involved at ALL :wink:

I'm getting an Ultimaker 2, if this guy comes through on it... should have it in a couple weeks! highest resolution and fastest FDM out there, as far as I know-- Open source, Bowden extruder (lighter, therefore faster and more precise), all metal frame... just waiting for him to pack it up. He had no luck locally on Craigslist so I talked him down to a fraction of the retail price, and he says he's only had it for 3 months-- very limited hours, looks basically brand new in the pictures! 20 micron layer height and really good XY resolution too, I think 100 microns? I'd have to check. 300mm/s extrusion, which is crazy fast. Obviously those a max parameters and can cause problems with some filaments... He's also sending it with a full roll of Woodfill, which I might use for dash pieces. Looks kind of like MDF when it's printed.

Awesome. That's exactly what i wanted to use! That's generous of you, and I'll take you up on it at some point!

That is an absolutely brilliant way to use a 3 speed hub! I like it! Shlumpfs are nice but like you said, super expensive. That's another reason I'm going serial instead of parallel, I just won't have to worry about gearing or chains at all, and that is nice. I'm thinking about putting my Cro in a larger wheel, as I'd like more top end-- to cruise at 50mph without OVS would be great. But I'm not sure what that would do to efficiency with the Adaptto. I suppose, realistically, I don't *need* to go over 40mph. But if I have a giant pack, and good aero, it'd be very nice to be able to do that without field-weakening. I really want to stick with a DD hub though, as the silence and lack of maintenance are keystones here...

Did you ever look at the Stokemonkey? Kind of what you're after, but more of a drop-in solution... and I suppose, not quite as powerful. Who am I kidding, you want to built it all yourself :D

I'll get some pictures together at some point, I have a bunch of scattered designs in OneNote and the self-centering geometry is totally stolen from Henry Thomas's Jetrike. I haven't decided which shocks I'm going to get yet but I think I'm going air, at least in the back where I'll have very low unsprung weight. As far as geometry, leading links are pretty typical. The swingarms will be suspended at their linkage to the rocker arm, and so will have a lot of travel.

Thanks AW, it's actually pretty easy to change the curves but I like the look too and don't think I will! =) That lotus looks classy!

UPDATE: Kink in those 3d printer plans-- the guy never shipped it, and I just got my money refunded via Paypal. Oh well. I still want an Ultimaker 2, due to its resolution and comparative affordability. Kind of wish I jumped on the Gigabot bandwagon when I had the chance, as that 2'x2'x2' print area would be very nice for custom fenders, fairing and long NACA cooling ducts, less piecing together. Which, for harder-to-bond things like fenders, opens up the game to more durable materials like nylon. ABS is super easy to join, but nylon is like Teflon. I'd need an ultrasonic welder. With a Gigabot, I could print single-piece wheel covers with radial NACA-ducted vents to my motor venting holes, to aerodynamically funnel air through it. Anyway, the price for a Gigabot went up $3500 to $6,000 after the kickstarter! I'm probably going to work on getting a nice welding table set up next, and then start on the frame...
 
After doing some simulations with swbluto's calculator, I think I may ditch my Cromotor and get a MidMonster, believe it or not. Based on measured values, it's predicted efficiency is greater than predicted efficiency with my Cromotor especially cruising at lower speeds (15-30MPH, in which it ranges from 86-92% efficiency) and my experience verifies that-- wh/mi are basically spot on, hard to do better than 25wh/mi without some sweat... but 35wh/mi is zipping around town worry-free.

Anyway, looks like I could do about 30mph on about 20wh/mi and get over 90% efficiency at that speed. And that'd be only 600 watts average draw... and those figures estimate my fairing would be 30% draggier than I believe I can build it. With the pack I want, I'd get 370+ miles on a charge with no human input... add solar and 100W of my own power, and theoretically an even longer trip is possible.

I've assembled a wishlist of welding accessories, but if my boss gets a 3d-printer which I would have access to, and it looks as though that'll happen, then I may build a tiny prototype frame first and experiment with changes before I even start buying steel stock. It'll be a CAD Christmas. =)
 
I considered that, but I want to build my own fairing around a delta frame because I want this bike to handle like a linear-track vehicle, but with more stability to facilitate winter riding and tighter cornering characteristics. The more I've ebiked, the more I've realized that freedom and a good view of the world is ultimately what I truly enjoy about it, and my current bike won't take me far enough, or through anything-- and it's not quite up to my comfort specs, though it's good. Nobody offers a leaning delta with sensible geometry. It's a niche. Also, deltas are excellent choices for cargo because of the undisrupted rear bay. Beyond that, I'm above average height and I want to ergonomically optimize this bike from the ground up. While I'm doing that, why don't I go ahead and add a few hundred watts of solar panels to that shell, I'm building it from scratch anyway, it's just money! As little as 200W peak, half of what I believe I can fit, would give me 600wh/day average-- quite a bit more during peak seasons. Since I am going to have this complex bulk/balance charging system and I want a FWD velomobile, why don't I just go ahead an do a serial-hybrid drivetrain instead? That'll be a tiny bit quieter, a bit less maintenance, and allow me to stay warm even when I'm not moving in the winter. And get exercise even if I'm riding at 100mph. And while I'm building this custom fitted frame and custom fairing, why don't I integrate a nice, custom-shaped 18650 pack in a subframe under all that too?

This hobby tends to get away from you, but as eager as I am to finish this project even before I've really started, I look forward to the learning experiences and all the well-finished integrated components I'll have when I'm done. Aerodynamics make normally impossible criteria possible if you spend enough money, e.g. 100mph top speed, 150-400 mile range + solar input, 1-hour charges, fully faired from the elements, ideal rider-specific geometry...

You get the point :p

[strike]UPDATE: Carbon fiber is awesome. BUT, I do need an anisotropic material for this project so I don't get a terrible benefit from it there, and it would require at least $1,000 in equipment just for proper layup, and then it either works, or it doesn't. Furthermore, UV can weaken it over time-- many consider 10 years a maximum usable life-- and I plan on leaving this in the sun on purpose. And finally, but perhaps most importantly, it's not the best thing to be surrounded in if you're in an accident. The splinters can be outright deadly, though I'd rather not think about that. I thought about an inner layer such as Kevlar, but after deliberating, I realized I could do something that's probably even more work, probably about as expensive, about as good, but way higher on the badassery scale and a useful skill to have, period... and one that would allow me to make complex and strong parts out of literal scrap metal with precision in the low tens of microns.

Investment casting. I mentioned it earlier-- basically, I'd be 3d printing a hollow shell of a part-- with a sprue-- and then dipping that in some ceramic slurry until I build up 1/2" over my print. Since I can join my prints for this process, I can make molds fairly large despite printing volume. I wouldn't need a huge forge because I wouldn't have a whole terrible lot of volume, but the examples I've seen are remarkably detailed-- 20-micron steps are distinctly defined in the final casting, from the original mold, in macro. Since I'd have "free complexity" from the printing process with these pieces, I could generate some sturdy and lightweight ribbed designs in Solidworks. With PVA and HIPS, this would be even easier than wax, as I would simply and cleanly dissolve out the mold for the casting. Futhermore, since these printed molds could be almost completely hollow, the total printing costs for the fairing would be less than if I printed ABS honeycomb-core pieces to wrap up in carbon fiber. They'd have to be layered with ceramic slurry, carefully dried, and then embedded in sand for the casting.

I'll look into it more...

PSS: Apparently aluminum thin wall sections can, under ideal conditions, be as thin as .2mm, while 1-4mm walls are typical. I might shoot for 3 or 4. Air bubbles can be an issue and commercial places use vacuums in order to force the plaster around the investment; this can get your precision into the sub-micron range! In my case though, if I had an aberration, I believe I could simply grind it off. Unless it came out really bad. I think I can weld sheets of that thickness with some toys for my MIG but it'll take practice!

PSSS: Doing the fairing with the investment method would definitely be more work, and probably not cheaper. However, it would be extremely cool to exploit the 'free' complexity of 3d-printing with casting to make a stronger and safer structure-- both fairings, fairing supports, and perhaps weight-sensitive parts like the swingarms. Not strictly necessary but I wouldn't want to spend ~2k and many hours on a nice composite fairing and have the sun weaken it to failure within a decade. Better to build one for life. By the looks of it, the ceramic slurries could be a great choice-- they would make it possible to cast, although with significant difficulty, an elliptical swingarm with a core pattern such as this: http://imgur.com/a/EyBXu[/strike]

So, I believe that I will stick to carbon fiber for the fairing itself, as I found some reportedly very effective UV-treatments that should extend the life of the fairing to beyond that of the solar panels, which is a reasonable lifetime. But I may utilize some small-scale investment casting to make mounts for the fairing.

Also, I found a vacuum pump that's apparently quite reliable when running for days, very well rated and reviewed, yet under $200. So that makes the composite fairing a lot less expensive! I found a 1/8th HP one, just big enough, for $700 from a composite supplier and $56 used on eBay under its actual brand name. What a racket. It could be used as a compressor too-- so I could, with a lot of work, blow a Lexan bubble down into a female mold for the windshield.
 
Since you're looking for max efficiency, you might want to run your numbers on HubMonster before deciding, because it's much lower phase-to-phase resistance means even less heat, especially if you're going to gear for significant speed and keep voltage below 100V. Kv=18.3rpm/v no-load 1.52A at 1000rpm and phase to phase resistance of only .016ohms . You'll probably find the extra 4kg or so worthwhile in nearly as compact a form as MidMonster at only 14mm greater diameter without the rim flange.
 
After looking through the Jetrike site some years ago, I left it with the impression that I don't want a leaner that keeps the CG at the same height, because I don't think it will feel natural to force the lean instead of initiating it with counter-steer just like a bike. I really only want the 3rd wheel for the extra traction, to permit lean lock when stopped, and to avoid a really long wheelbase while getting the CG very low. I think I'd shoot for about half way between neutral and falling to the side the same as a bike. Are there any handling negatives with that, because if I build one it will be for rocketing on the highway and carving up mountain roads?
 
John in CR said:
After looking through the Jetrike site some years ago, I left it with the impression that I don't want a leaner that keeps the CG at the same height, because I don't think it will feel natural to force the lean instead of initiating it with counter-steer just like a bike. I really only want the 3rd wheel for the extra traction, to permit lean lock when stopped, and to avoid a really long wheelbase while getting the CG very low. I think I'd shoot for about half way between neutral and falling to the side the same as a bike. Are there any handling negatives with that, because if I build one it will be for rocketing on the highway and carving up mountain roads?

I am considering HubMonster; the efficiency figures at lower speeds are good: I calculated over 86% at 25mph in a 19" wheel at 90v. But what's really great, is that I should be able to hit 65mph with less than 3kw at 93.5% efficiency on the flat. Diameter isn't an issue as I'll be building the forks up myself (and they will be extremely beefy triple-clamp leading-link forks!)

Fun tidbit; it appears by my math that the ideal OD for efficiency at top speed with Hubmonster is 18.9" at my 90v. Golidlocks hubmotor!

You'd still counter-steer like on a regular bike, it'd just take a little bit more counter-steer. The best part is that I could pretty easily weld up new swingarms and/or pivots to adjust track width and CoG offset if I don't like how it turns out. I doubt the effect would be enough to remain upright but that's hard to simulate so I'll have to wait and see. It would just result in a more stable bike, although that's not as necessary with a FWD design anyway =) There are many examples of leaning deltas with geometry that has the opposite effect, and anecdotal reports indicate that they take some more practice to ride and result in an extremely responsive bike. I don't plan on riding technical obstacles (and deltas aren't ideal for offroading anyway with 3 tracks) so I will try the more stable geometry first.

My lab has one week to spend the last bit of their last RO1 grant and I'm currently trying to find the best way to spend $10,000 on 3d-printers. FDM really is the growing king, surprisingly... they're just too cheap to be true. 20-micron prints look absolutely crazy. Here are a couple examples:

Yoda with 20 micron layers-- that's .00079 inches per layer;
6191494335_02174b374f_b.jpg


And it's a bit big to post here but he did another print with 20-micron layers and put it under a microscope. The blobs on the left of this picture is the guys fingerprint...! https://i0.wp.com/www.hive76.org/wp...hFingerAndScaleBar_162pixelsPerMillimeter.jpg

Absolutely begging to be used for casting molds. Especially with PVA/HIPS; I believe with HIPS you could just use it as you would a hand-carved polystyrene mold but I haven't seen it done. Basically, you don't have to burn out your mold from your investment; it just burns off into vapor when you pour your metal!

I suspect with a finely tuned printer, foundry supplies, and a drill press I could pretty much avoid the machine shop entirely.
 
xenodius said:
Fun tidbit; it appears by my math that the ideal OD for efficiency at top speed with Hubmonster is 18.9" at my 90v. Golidlocks hubmotor

With the stock 13" scooter rim, you can probably get to that with a 13" tire for a front instead of the wide rear I use, which ends up a 19.25" OD wheel running the lowest profile tire available.

When you're running numbers, keep in mind that the measured resistance was at 25°C, so even with ventilation heat loss in the copper is likely to be 15-20% higher at operating temps unless it's chilly outside.
 
John in CR said:
xenodius said:
Fun tidbit; it appears by my math that the ideal OD for efficiency at top speed with Hubmonster is 18.9" at my 90v. Golidlocks hubmotor

With the stock 13" scooter rim, you can probably get to that with a 13" tire for a front instead of the wide rear I use, which ends up a 19.25" OD wheel running the lowest profile tire available.

When you're running numbers, keep in mind that the measured resistance was at 25°C, so even with ventilation heat loss in the copper is likely to be 15-20% higher at operating temps unless it's chilly outside.

Yeah I know! I was actually pretty happy about that. And they were calculated like that, though in real life winding temps will be yet higher with some stop and go riding. Still should be the average for long cruises, and with a giant pack I do plan to take long cruises. Even higher discharge rate 18650's are just ridiculously great energy density compared to my pack... :lol: nuts.

Adaptto is soon releasing an integrated DC-DC converter that will make this whole build a bit cleaner! =)

EDIT 12-21: Waiting to hear back from Oleg about my current Max-E, I asked him some questions about charging. I'd like to be able to hit at least 7kw charging input, and it looks like that'd be doable really without any mods if I had a 6-phase motor, especially one with really low resistance. I've been looking at a lot of simulations and I really believe that Hubmonster will still be very efficient at lower speeds, even down to 20mph it beats any >3kw hubmotor I can find. I should get well over 80% efficiency going up a 10% grade at 25mph, and I should really never have to go less than 25mph. I think part of the reason I had so much trouble sipping 500-1000W on my KMX, is that it's an aerodynamic parasail. Compared to a faired bike it has so much drag that I couldn't keep the RPM's high enough. Shoudn't be as much of a problem with this bike. It'll depend a LOT on rolling resistance, though! Still, if I actually get 85% efficiency and 30mph on 500 watts I'll be so happy...

I should be able to do 100+mph with OVS, too! Dunno if I even want to join that club... but I might not be able to help it with that power at hand.

I'm looking at arduino-controlled LED-drivers and considering making my own turn-signals, integrated with a 3-input (RGB) LED strip along the entire length of my fairing, so that signals blink the corresponding side yellow on top of headlight and rear blinkers in my rear wheel fairings. Braking would trigger red on both sides, and hell, I could even tap into my throttle hall and make it light up green when I hit the gas. Plus, generate a multitude of patterns for even more nighttime visibility.

I'm getting really excited to get rolling-- I missed out on a great 10'x5' and 1.25" steel-top workbench last week, because it took too long to make arrangements to pick it up-- I'm sure it was well over a ton, and on some crazy casters too. Said it was rated for 2 tons. So overkill, but he only asked for $100! Worth more in a scrapyard. If I can't find a good bench on Craigslist, I'll break down and buy a welding table and build a wooden bench for everything else. And I've settled on a Gast pump for sure, I'll use it for both blowing lexan and for vacuuming carbon-fiber.

Should have some time over the next few days to do some modeling, hopefully. There's tons to do because I want to get my sketches for the frame into Solidworks, and I'll build it a few different ways. It'd be awesome to print and piece together a scale model too. I'll need it all in solidworks eventually.

Excited for what's possible...
 
So, since my lab does optogenetics I have access to a really nice Thor illuminance meter. I used it and a semi-flexible monocrystalline cell for a preliminary experiment.

The results were fairly disappointing. While I really only need about 300wh/day for most usage, I was excited at the prospect of netting ~1kwhr during summer days which would have been possible based on my math. However, it appears that more than geometry is at play when panel efficiency and varying angles-of-incidence are at hand... Looks like real-world efficiency is as low as 20% of what you'd expect based on simple geometry*. This also means that by mounting panels along a curved fairing, I'd be getting ~30% rated efficiency at best. That could be increased to ~55% if I used my own cells and did a ridiculous amount of custom wiring to keep those lagging cells from dragging the module down, but unfortunately, that's too much d*** time, work, and money. To do this properly, I'd have to take my final fairing design, print a plug so that I could inject foam into a form barely larger than my intended fairing, and then carefully remove material to manually "tile" the entire form to fit individual thin-film solar cells without actually flexing them and wire them together with inline diodes for each cell. Oh yeah, and I'd need an MPPT too-- and I'd really have to build and design my own from scratch to make it all worth it.

TL;DR: When semi-flex panels are bent, they are less efficient than simple geometry suggests, and I won't use any panels due to the incredible added complexity required for proper implementation. Commercial modules and commercial MPPT's are out.

*e.g. cos(60 degrees) from ideal, you'd expect 1/2 of rated power. That wouldn't be shabby if it were true.

The good news is that this simplifies the whole process a bit. Now it's just CAD, carbon, welding, and wiring.

Oh yeah, this weekends project; my first all-grain batch of homebrew. If it turns out I'll share some with you, Brake. =) It's a Biere de Garde, shooting for high lovibond, carb and gravity for the style, but low dextrins (light bodied), and I'm doing my secondary fermentation on oak chips soaked in a particularly minerally unoaked Chardonnay from Mersault; Patrick Javilier. I hate it by itself but it's fantastic paired with... almost anything, and i have a bottle to burn. I think it's the perfect French wine for a French farmhouse style. Should be dark, but only with toffee/caramel malt character, light and bubbly, a hint of spicy phenolics and a myriad of fruit (figs/dates from yeast; apples to stonefruit from the wine) all the way through some subtle oak. I hope. :mrgreen:

7Sj3kmcl.jpg
 
Congratulations on the first batch! I'm sure it will turn out crazy good. I still don't know how you can fit all this into your schedule. :)
 
It'll be good as long as I don't mess up on something-- a lot can go wrong with all-grain. But I'm controlling for everything as best I can. If I'm jaded then it'll be harder to be disappointed =)

Grad school is no problem, just have to do everything really fast :mrgreen:

Cleared a space for a 3d printer today; new boss definitely wants one and we talked about options a bit. He seems set on it so I'm not looking for one now, he's the kind of guy who'd encourage me to use it, so I can save more money for that giant battery! And I'd sure use it-- right now I want to print a custom-fit housing for my in-ear monitor armatures that our puppy chewed up, replacement headband for my headphones, handles for the fridge, LOTS of pegboard accessories, and then odds and ends-- like a soda rack for our fridge and a butter pig.
 
I generated some CFD simulations, finally-- I actually lost count of the number of times I had to reinstall Solidworks. And it's actually pretty picky about the parameters of the model too... here's what I got, this is a pretty poor model and the numbers (94.3 newtons at 88mph) are a bit higher than my expected CdA, but should be more accurate once I find a way to use guide curves without breaking my lofts... not to mention, no windscreen on this model. Just the uncut fairing. So the headrest is like pulling a wall instead of smooth curvature.

http://imgur.com/bOsS7ua,60gamwz,pPm8CSY,DkAqnfh,9UE5YoO,D4PWh6J,ycpjMRv,OvSwRBW#6

What I really wanted most of all;
OvSwRBWl.png


And here, a telling indicator of inaccuracies (and non-laminar flow) resulting from Solidworks preventing me from utilizing lengthwise guide curvature... not sure why yet. There are great tutorials out there though so it's just a matter of time. I also need to extend my tailbox, as you can see there's a pretty substantial pressure zone around there.
60gamwzl.png


Oh yeah, and if you're wondering why the lower-right corner looks funny, it's because that's cupped for the front wheel. This shot really shows off how bumpy this model is... :? Just need to understand the software better.

bOsS7ual.png


And a cut plot proves useful:

nLpVYpAl.png


I think the tailbox should extend at least to the edge of the blue low-velocity zone-- it's also pretty low-pressure.
 
That's a little wierd about the modelling problems. When I first saw it before reading the post itself, I actually thought you'd modelled a crashed version. ;)
 
amberwolf said:
That's a little wierd about the modelling problems. When I first saw it before reading the post itself, I actually thought you'd modelled a crashed version. ;)

Seriously. Did some reading about requirements for surface lofts and no answers. I sketched a bunch of vertical ribs, and then 3 guide curves-- top, bottom, and one along the outermost edge. Unfortunately, it simply 'failed' when lofting across the ribs, guided by the vertical curves. But lofting along the horizontal guide curves using the vertical ribs as guides works. Sometimes extremely small issues can prevent something from working so I'll probably just start a new sketch, but perhaps once I finish the frame model; that's a little more bite-sized.

90mph winds better not deform the fairing that much =) I don't think Solidworks has a real FEA material for carbon-fiber, and it varies so much with each layup it'd be no surprise. But I'll overbuild it a bit so bystanders don't push through it on accident anyway...

I've been very eager to use the CFD on a somewhat-realistic fairing primarily to see if my expected ranges are really in the ballpark. And I think they are... so I'll focus on the frame until I get that done. If 200+ mile range wasn't feasible, then I'd have to move to higher-power and bigger packs, which means 1 hour charging wouldn't be feasible, which means long trips for no reason wouldn't be feasible... you get the idea. Some of the practicality of this design hinges on efficiency. I'll be taking short trips and occasionally, very long ones.

Much easier to work with solid bodies, the swingarm only took about 15 minutes!
OvOYO4i.png


I hope to print and assemble a miniature model first.


EDIT: I am stealing VeloTilt's swingarm design--it's brilliantbecause it allows you to quickly swap out the swingarms with a single bolt.

The pictures illustrate it best, but they've basically used a piece of diagonally-cut stainless tubing welded to tubing on one half, and polished/tapped rod on the other, with a bushing between. So, they can run a bolt right through one side, and pull the tubing flush together, but support it all by the bushing. This would make it an absolute breeze to swap out wider, 'winter' swingarms.

EDIT2: So my lab ordered an Ultimaker 2 today, or at least put in a request to order it. It's going to be my pleasure to calibrate it and make any parts anyone needs... work out the kinks. I have a long list of things to print, some essential and some for fun... oh yeah, and some actually for the lab =) I'll probably not get anything done on this build until summer, except a couple days around spring break. Still need to do some design and FEA before I make a cut list, and get all our boxed up junk in the garage onto shelving to make room for a proper workbench... currently my improvised workspace is a bit small and short...

I'll use the printer to make 18650 spacers and, after looking at the Rotovelo's rigid rotomolded fairing, I'll probably try a prototype ABS fairing before I commit with carbon. Too cheap not to test the design first, at least relative to carbon. Brake has built an incredibly nice spot welder that I'll twist his arm to use =) Maybe throw in some beer. 440x 29E cells arriving in a month or two, I'll be buying more when I have a couple thou to commit but those cells as they are should be good for at least 12kw ouput, so I could build a partial pack from them and still have 4.8kwhr as is!
 
Great post......
thanks Xenodius!
I fully support your 'statement of directions' 8)
Personally I have the same vision for everything you mentioned with just one exception.
I don't want several KWh on board, not enough 'sporty' for me.
The beauty of this solution should be of course 'long range' (200/300 miles) but with active contribution from the pilot still pushing on the pedals.
Otherwise the word 'VELO' doesn't make a lot of sense in this project :lol:
By the way here in EU the legal limits are 250W, which makes everything so difficult.

Anyway all the best and success!
 
Conforming to your lower wattage limit means that a more typical velomobile and a mid-drive are almost certainly the best solution =)

My goal is to keep the finished bike 200lbs, the pack will be at least 100lbs and about 10kwhr, charging time will be at least 1.5 hours I think. I have about half the pack on its way right now =)

I think a 10kwhr pack might go well over 400 miles on a mid-drive at a good clip. I don't expect it to accelerate quite as fast as the bike I have now (at least in the 0-30mph range) with the extra 100lbs and FWD instead of RWD, but the motor is capable of twice the power (I'm going to run it 24-28kw) up to twice the speed (88mph vs 43mph) and that's plenty for me. Keys here are cargo, comfort, and range... tilting facilitates cargo, and comfort. I also want a narrow enough track that I don't feel out of place if I want to mosey along at 15mph on a MUP.

Sidestepping here. Jetrikes geometry requires that my CoG stay low to facilitate the self-centering effect. Larger wheels and certain geometry modifications can extend this beyond the 'rated' 1/2 meter, but having that large pack mounted low will actually really help me implement the fairing the way I want to-- with only tiny round holes for the front fork, and rear swingarms... and a drain. And have enough room to meet that zero-point turn criteria.

I suspect that the battery pack alone might be enough to keep the bike from tipping over when parked. If not, then that's what the lean-lock is for.
 
xenodius said:
10kwhr, charging time will be at least 1.5 hours I think.

Heckuva charger! I have a 2-space full-width rackmount Sorenson lab PSU that's ~55v @ 55A, and to get the full output of around 3kW it requires 3-phase 220VAC input to it.... With only 2-phase it's limited to something like 1.5-2kW, I think.
 
Back
Top