Honda Insight Forum banner
Cancel
Create thread New Garage Forums

Stacked Insight IMA Motors

Jump to Latest
45K views 128 replies 23 participants last post by  Hugh-Falls  
H
#1 ·
The idea of stacking two or more Honda Insight IMA motors together in tandem fashion has been suggested by others for powering an EV or for enhancing the Insight's battery only performance. The ORNL study on the Accord IMA motor shows what might be expected and a way in which this type of adaptation may be accomplished. We should be able to make up some end plates, adapters, bearings, and shaft for an Insight setup.

An Insight specific conversion/adaptation probably should 1.) bolt up to existing hardware in a tandem arrangement and/or 2.) use Insight parts in a free standing motor for general EV propulsion. A lot of work has been done by Mike and others on the electronic end and hopefully we should be able to work up some practical motor hardware also.

To begin with, on the Insight ICE engine, can someone furnish the dimension between the end of the ICE crankshaft (IMA motor rotor mounting surface) and the machined surface on the ICE block where the IMA motor is mounted? In other words, how far beyond the engine block does the crankshaft protrude (or not). With this dimension we should be able to work out the critical relationship between IMA rotor and stator.

Hugh-Falls
 
#3 ·
I think two IMA motors stacked would give 10kw continuous power with a peak of perhaps 30kw if over driven. Enough for a slowish EV and max speed of perhaps 60-70mph.

Three would be better giving 45kw peak and probably 80mph+ top speed

Four would be the performance version giving 60kw and getting on for 100mph I bet.

Four stacked would be about around the width of the IC engine + IMA motor so would fit in the standard space. If they were aligned carefully you probably would only need one set of commutation sensors and could drive them in parallel with a single highpower AC drive say the siemens unit or similar.

Uping the voltage to 200V or so would also boost power.

If you remove the fuel tank and all the other stuff including all the old IMA gubbins you could probably fit a decent battery pack in the IMA compartment possibly split and have some underneath in a box as well.

I don't think it's practicable to use the existing IMA drive to run multiple motors, although it might just run two but would need lots of reverse engineering.
 
H
#4 ·
retepsnikrep said:
..............
Four would be the performance version giving 60kw and getting on for 100mph I bet.

Four stacked would be about around the width of the IC engine + IMA motor so would fit in the standard space. If they were aligned carefully you probably would only need one set of commutation sensors and could drive them in parallel with a single highpower AC drive say the siemens unit or similar..............
Click to expand...
It appears that what we need is a free standing triple stack motor with a mounting face identical to the ICE (which it will replace) to which the existing IMA motor may be mounted and synchonized so that the existing commutation system may be used for all four. In addition, the power bus will have to be extended to include all four motors. The entire assembly should be totaly enclosed because the strong magnets used could accumulate debris over time. Simple provision for water cooling the motor(s) using existing plumbing could also furnish cabin heating when needed. The use of a commercially available motor control unit would remove a lot of uncertainty from the project.

Hugh-Falls
 
#6 · (Edited)
I have looked at this issue for some time, and have 4 of the IMA motors, as well as 15 or more civic power inverters.
The magnetic field issues that Peter and I were initially concerned with do not seem to be a major issue.
I base this on a test of the magnet strength inside the coil assembly, and outside. When the magnets are on the bench, they will yank a hammer out of your hand. When they are in the coil assembly, a screwdriver barely sticks directly on the magnet. This is due to the intense magnetic flux path and narrow gap between the coil structure and the magnets.

I was thinking that one IMA motor for each rear wheel would be a better way to go than to make one motor with two rotors. A single motor on one wheel would probably be all any hypermiler would need.

Sink the flat IMA motor into the body so the motor output shaft lines up with the rear suspension hinge point, and weld a large timing belt sprocket to the rear wheel.
Pick a sprocket ratio that will let the drive do 90 mph at the motors max rpm.

Have a switchover system so the stock IMA motor can be switched to the rear motor, and the stock drive electronics could give an EV mode.(MIMA)
As far as max power from a permanent magnet motor, I am no expert, but understand that the limiting factor is demagnetization of the rotor magnets, as well as how cool you can keep the coils.
Series wound motors typically used on home built EV's can do intermittent burst of 20 or more times the average power, where PM motors need to stay closer to their rated output.

The Prius MG2 is a pretty nice drive motor if more power is desired, it has bearings, sealed case, cooling and a max rating of ~50KW.

The interesting thing about the honda and toyota motors is that the rotors all have about the same diameter.
The more powerful the motor, the longer the magnet rotor assembly, MG2 is about 8 inches deep, compared to the toyota MG1 and the IMA motors ~2.5 "
 
H
#11 ·
Mike Dabrowski 2000 said:
Guess I don,t understand why you need someone else to give you that dimension, if you have the components on the bench, all the pieces required to make the measurements are in front of you, or am I missing something?
Click to expand...
Mike: All the rotating parts between engine and transmission have a starting surface somewhere and I chose the end face of the crankshaft because it is where the first part I have (rotor) locates. All the non-moving parts between engine and transmission have a starting surface somewhere and I chose the mounting surface of IMA motor frame at engine block because that is the first non-moving part I have. I could speculate about the axial relationship between rotor and stator and could probably be wrong. The dimension between the end of crankshaft and the IMA housing mounting face will no doubt vary from car to car due to manufacturing tolerances but will be an actual functional starting point for a layout.

This dimension could turn out to be the same as the one with rotor at rest placed within the stator on the bench and maybe not. Hopefully, someone out there has an engine available for measuring without the IMA motor attached.

Hugh-Falls
 
#12 ·
Hugh

I had a look at my spare engine today which still has the IMA motor and rotor attached. I don't have the special rotor puller and don't want to disasemble it any further at present without it, so I'm sorry i can't supply the measurements you were after.

Perhaps someone with the ima seperated from the engine can oblige.
 
#13 ·
Hpw hard is it to remove the IMA motor. I have considered pulling the one off of our parts engine.
 
Save
Reply Quote Like Like
H
#15 ·
It has been suggested by others elsewhere that when the six bolts securing the rotor to the crankshaft are progressively loosened and removed, the rotor will snap in some random direction against the stator and remain there after the bolts are removed. It has also been suggested by others that several layers of heavy stiff Kraft-like paper be inserted around the rotor in the air gap to cushion the fiber glass rotor band as it snaps against the stator. Installation is the reverse process of progressively tightening the bolts to bring rotor back into centered alignment. It is suggested that the paper shim around the rotor extend wider than the rotor so that it may be gripped for easier removal in case it gets imbedded in the stator laminations. I don't know if any of these methods work but they sound reasonable, especially if there is no reason to have to separate the rotor from the stator. I do have some concerns about the locating pin, however.
 
#14 ·
A procedure that I have used:

The rotor will jump to the coil assembly once the rotor bolts are removed from the crank.
Once that happens it is difficult to separate the rotor from the coil without damaging the Kevlar filaments that surround the magnets.

If you slide some brass shim stock, manila folder or other spacer material between the coils and the rotor, the material will protect the filaments.
The coil and rotor can be removed as a unit at that point.

If you want to separate the rotor from the coils, place the rotor and coil assembly on a gallon paint can, can centered on rotor and press down firmly on the rotor as the paint can lifts the rotor out of the coil.

The protective material will abrade on the coil laminates, but will protect the filaments.
To re assemble, put the coil on a table, paint can in the center, rotor on top of the can, shim material wrapped around the rotor,and protruding into the coil as a guide and lift the coil assembly. Don't get your fingers between the rotor and the coil assembly or you could loose the ends, as the coil will jump up tothe center of the coil with a lot of force once the rotor gets near the magnets, but the shimstock will prevent damage.
The rotor and coil assembly can then be bolted to the block, and the rotor will recenter as you get the bolts started. Make sure that the rotor is positioned so the locating stud on the crank is in the correct position or you will have to rotate the crank, as the rotor cannot rotate once it is stuck to the coils.
When the rotor and coil assembly is on the bench, the mounting surfaces that you are asking about should be easy to accurately measure.
Put a straightedge across the coil rear mounting surface, and make sure that the rotor sits centered front to back in the coils, and using a digital caliper measure the difference. I expect that the front to back variability car to car can be more than .03 and the motor will work just fine.
 
#16 · (Edited)
I used brass shimstock with good sucess and Ron Hansen has used manila folder material.
The redlight racing guys also used this procedure to mount the IMA to their TDI diesel insight.

If you only want to make your measurements,the relationship of the rotor to the coils and crank will not change if you don't separate them, so you should be able to simply put it back on and the pin should be in the correct position.
I have been asked this question and have given this advice to at least 10 people, and no one has told me that it did not work for them.
 
#18 ·
I've finally got my four IMA motors on the bench now, so need to take some very careful measurements. I like the idea of a 40-80kw unit i can pick up on my own :) Thanks to Ron & Trent in USA for help with sourcing and carriage.

Any ideas and thoughts on joining them gratefully recieved.

Looking at them with some slight machining of the housing the stator units will certainly fit together, but how to connect the rotors to each other and support them is the issue.
 

Attachments

#19 · (Edited)
That's a nice picture Peter.
You are looking at a lot of man hours of work.
The way I thought would be the easiest would be to make aluminum spacers that would fit between each rotor that was exactly the thickness required to place the rotors in sync and in line with the coil assembly's.
The spacers would have through holes so long threaded rods could hold them together.
The encoder disk would be on one end of the rotor stack, and an output shaft would bolt on each end of the stacked rotors.
If all the surfaces were milled accurately, and the rotors pinned, you would have a rotor assembly that was stacked to any amount of rotors and coil assemblies, and would only have to have two types of components.
The first would be a welded assembly of a flange and output/input shaft, and both would be the same.
The second would be the spacers with holes and locating pins, and you would need three identical spacers for your 4 motor assembly.
No machining of the rotors or housings required.
Each rotor could be sitting in the coils assembly it will rotate in, and the spacers placed on each motor. when the whole assembly is bolted through with the long threaded rods, and the output shafts mounted on a bearing plate on each end, and you have the worlds first Quad IMA drive.
A lot of fairly precise machining, but not something that most machine shops would have problems with.

Good luck.
I have 5 IMA motors sitting on my what do I do with these shelf right next to 10 Civic power inverters.Hmmmm!
Will be interesting to see if you can get to that project with all of the other cool stuff you are playing with.
 
#21 ·
If the motor is hotter than the block, heat flows out of the motor. The block is cooled by the antifreeze.
The small fins on the rotor and the tight proximity of the surfaces and a small fresh air inlet make the spinning IMA motor circulate air.
I would call that air cooled with a warm block as a passive conductive cooler.
Neither is a very effective cooler, so the idea of putting in some liquid cooling like Chris did on his dual motor setup is probably a good idea on a quad setup.
 
H
#22 · (Edited)
Quad Stack

We all know that the whole is something more than the sum of its parts. At present my stacked IMA motor is a bunch of parts and "there are many slips between the cup and the lip", so no "in progress photos" of parts.

1.) If the power cable lug protrusion is cut off and the area leveled, the dowels will key the cases together nicely and the bolt holes will also line up for through rods to hold the assembly together.

2.) A flanged full length keyed shaft mounts the Insight flywheel/sensor/clutch assembly, rotors, and bearings. In my case, this shaft is machined from a Chevy '81 pick-up rear axle which is an excellent flanged steel forging. Other axles should also be adequate.

3.)The rotors are mounted to the 1-3/16 diameter keyed shaft using standard industrial power transmission components. Adapters are machined from a standard #35 pitch sprocket with "#1610 taper bore bushing". The adapters fit the flywheel pilot diameter of the rotor and use the same bolt holes and indexing dowel. The ID of the rotors are enlarged to accept the 1-3/16 diameter shaft: this requires a few swipes of a file on each of the six crankshaft mounting screw lugs. A drill jig was made using a #60 sprocket with "split taper bushing (has internal and external keys)" to drill the dowel and bolt holes in adapters to insure easy accurate reproducibility. The #35 sprockets were mounted on 4" long keyed stub shafts for machining and subsequent drilling (the drill jig was keyed and mounted to each stub shaft for drilling).

4.) The stack, starting at flywheel assembly, consists of an empty case with intact power lugs and sensor assembly; a 1/2" bearing plate carrying the front bearing cartridge which supports the motor shaft and absorbs clutch release bearing thrust; four stacked stators/cases; and an end plate machined for stator clearance carrying the rear floating bearing cartridge and drilled for motor mount. At present, shaft extends beyond rear bearing.

5.) Spacers are required between the rotor assemblies to maintain correct stator/rotor relationships. Assembly will be attempted using Mike's shim method with heavy duty drafting Mylar sheet shims and a case lowering fixture, one layer at a time.

IF and WHEN the motor is assembled and works, photos will be posted.
 
#28 ·
Hugh-Falls said:
1.) If the power cable lug protrusion is cut off and the area leveled, the dowels will key the cases together nicely and the bolt holes will also line up for through rods to hold the assembly together.
Click to expand...
I noticed that which is nice.

Hugh I appreciate your reticence about posting photos but we are desperate to see the work in progress for ideas and inspiration. The chevy axle thing seems intersteing but unless i see a pic i have no idea what we are talking about.

I think there are three/four of us now actively looking at a stacked motor setup. Chris has done some great work and so have you by the sound of it. We are not wanting to judge or criticise what you have done. It's a tricky project and has several solutions i'm sure.

Having been looking at the stator attachment points inside the alloy case I doubt that transfers much heat direct to the case through conduction itself. Must be the circualtion or air in the case which moves heat from the stator to the alloy case. With that in mind I will probably go with a pressure air blower system pushing/sucking air through from say one end to the other. I'll have temp sensors on each stator and an infra red sensor on the rotor farthest from blower.

Mike like your ideas as well need to talk to my friendly machine shop.

Re the Insight EV conversion i think that's a missed opportunity, any conversion nowadays using DC series motor and lead acid is prettty old fashioned 30 year old technology. Huge weight penalty as well.
 
#23 ·
Wow, looking forward to seeing this.

Sounds like you're making some good progress. But I want to see pics even if it doesn't work. ;)
 
Save
Reply Quote Like Like
#24 ·
A bit of topic,
I just had a nice long conversation with a guy in Walton Kansas that has been building EV conversions since the 70's
EV-Blue.com, Electric Vehicle Questions and Answers, EV-Blue.com, EV Blue, Electric Blue, evblue, evblue.com

He is just finishing up an Insight .
He used a Netgain 9" brushed DC motor with 12- 12VAGM lead acid 27 batteries, and custom springs to carry the 450 lbs.

He asked me if I was interested in buying any of the Insight parts, he has the engine with IMA attached, all of the IMA stuff including the battery pack.
May be worth a trip to his place if you live in the area, and need any of the stuff?
 
M
#25 ·
Group 27 AGM's? This car begs for a lithium pack, 450 pounds and probably a 40 mile range after the peukert effect to dead if lucky and 15 or less if he wants the batteries to last and all the weight? I have respect for Wayne but putting lead into an Insight boggles my mind. It seems to make some sense for the lead sled pickup trucks where he can load 900 pounds of lead in the back, but for a small car? At least he used AGM's but they are going to be about 100Ah cells. The same weight could easily get 100+ miles of safe discharge range with a lithium pack and much nicer performance too.
 
#26 ·
Yeah, but probably at least 2x the cost...
 
Save
Reply Quote Like Like
H
#27 ·
Please redo the math

Gentlemen:
After going to the "Blue" website and reading your replies above, reaching some of those range/weight conclusions seem improbable with a 144V brush type DC motor installation.

For $12000 plus, you should hope to get more for your money than 1970's lift truck technology.
 
#30 · (Edited)
Based on my earlier experience with lithium EV's and the 40ah car and having pushed a full 40ah of capacity through a single ima motor in a little over an hour i never experienced any motor or mdm problems. Average current ~ 30A for 1hr 15mins average power say 4kw at 144v nominal.

I would estimate without engine drag that 6-8kw would provide enough power for 60mph crusing speed without problems. 4 motors will be at an average of around 15-20A each for 8kw of power. That's nothing in the grand scheme of things.

Short bursts of around ~20kw per motor could be expected with a suitable battery and some final tweaks to the MDM setup giving a pretty nippy 80kw peak power :)

Based on all this I do not have any undue concerns about cooling, and a simple air in one end, air out the other end, with a heater scroll type fan should be more than suffcient IMHO. Ducting this warm air for heater use would be sensible if it can be fed into the air intake somehow. I look forward to seeing the other solutions though.

There will also be some (not much) heat available from the 4 stacked MDM units which i will modifiying and mounting directly above each motor in some fashion, again if this heat can be ducted that will help. A ceramic heating element inserted in place of stock heater matrix will almost certainly be reqd IMO as heat generated will be pretty low.

Temp sensing is all part of the package of course :)
 
H
#32 ·
Insight EV heat management system

retepsnikrep said:
I would estimate without engine drag that 6-8kw would provide enough power for 60mph cruising speed without problems. 4 motors will be at an average of around 15-20A each for 8kw of power. That's nothing in the grand scheme of things.

There will also be some (not much) heat available from the 4 stacked MDM units which i will modifying and mounting directly above each motor in some fashion, again if this heat can be ducted that will help. A ceramic heating element inserted in place of stock heater matrix will almost certainly be reqd IMO as heat generated will be pretty low.

Temp sensing is all part of the package of course :)
Click to expand...
1.) Why not keep as much of Insight climate control plumbing as intact as possible and continue to use a water based heat transfer method? Keep radiator, heater and air conditioning as is. Add an electric water circulating pump and a drive motor for the AC pump.

2.) Air "in" at the stators and "out" through the rotors appears to be the simplest method for cooling the stator windings and the rotor magnets. Use a stripped Insight heater with blower, filter, and controls as a heat exchanger to recover motor heat. Air cooling would be a closed circuit.

3.) Fabricate 1" thick drilled aluminum water cooled heat sinks and mount an MDM on each side making it possible to mount two pairs of MDMs in an enclosure in the limited space atop the stack and recover the heat directly in the water. An additional wet heat sink is needed for DC/DC converter.

4. An external 12 volt block heater would be added to provide heat as needed. A plug-in adapter to a household power source could be used to preheat the system in cold weather.
 
#31 ·
EV-Blue, did indicate the lead was selected due to cost , but I had the same questions as to why lithium was not selected.
The guy has been doing a lot of conversions, 175 in 1980 and understands the electronics and batteries he has been using for the past 20 years, so of course he works with what he knows best, may not be optimum, but it still works.
He has another Insight to do in the near future, and will be in contact with me about the Synergy drive project that I really wish I had the time to work on.

While we would love for the stacked motor to not require any active cooling, I think that ChrisMA's approach to water cooling the stack is the way to go. Winter driving in an EV that has no heat is not much fun, so I would go through the trouble of putting in an antifreeze and water pumped system with the ability to put the generated heat into either a heater core for heating the cabin with fan or a small dump radiator for dumping the heat in summer. The same with the drive electronics. That is where the big heat is generated. make the air cooling system so the exhaust air can be vented in or out of the car. I got to see how much heat came out of the inverter when the WWU x-prize car was here. A one mile drive got the inverter heat sink so hot it would burn skin when it was operated without cooling air.
Keep at it.
 
#33 · (Edited)
I think the main problems are there won't be very much heat and it won't be very high quality i.e. not very hot. A lot of effort to recover perhaps 1kw of heat if running at 10kw load. To recover that meagre 1kw from motor and mdm and transfer it effectively to the wet system is just not practicable IMO. Moving the slightly warmed but more directly useful air about i see as a possibility with some cheap ducting. I'll be removing the aircon from my car anyway as i never use it.

The dc-dc converter can probably be mounted on a single heatsink exposed to the air flow at the front of the motor stack where throttle body and water radiator etc are now.

I have successfully used a 2kw US hair dryer in my van for heating. A fan heater matrix inserted into the air box can do the same thing and yes I used to preheat the car/van interior from the mains a few minutes before setting off.
 
M
#34 ·
About 30 minutes of a 1.5kw heater running in the back of an Insight is about all that is needed for -10f(-25c) temperatures to be comfortable when setting off to about 10 minutes into the trip where the amount of air required to keep the front glass from fogging up makes the interior cold. Of course with electric heating that air to the windsheild would be heated and it wouldn't need to be as warm when setting off. 15 minutes from the grid would more than take care of it for reasonable comfort if starting from a garaged area and the windsheild is already clear of fog and ice. If there is iced fog and snow/ice on the outside, 1.5kw would be far too little to remove it. Most US hair dryers are closer to 1500 watts because our 120 volt 15 amp bathroom circuits can't quite handle 2kw and some leeway is needed for other appliances like battery powered shavers and the like to charge. I've venture a stacked set of 3kw should do it, although I'd almost be inclined to have 4.5kw available to clear the windows for when parking outdoors and getting an ice glaze on the outside and frost inside. I'd rather ditch the costly electric heating and use a fuel for that, seems to suite the purpose better but safely integrating something like that in a car is something I don't know if I could figure out, I'll probably stick with electric heating.
 
Honda Insight Forum
posts
498K
members
190K
Since
2003
We’re the ultimate Honda Insight forum to talk about Honda’s hybrid car and its fuel economy and specs!
Full Forum Listing
Explore Our Forums
Honda Insight Forum 1st-Gen Discussion Modifications and Technical Issues Problems and Troubleshooting Member Classifieds Honda Insight Forum 2nd-Gen Discussion