[retepsnikrep]

Below a battery resting voltage of about ~140v the (MCM) IMA will not operate the start cycle. Assist is available down to about ~120v under load. The system current increases as voltage fall as long as cells are in balance as IMA trys to keep power constant with falling battery voltage. This load has to be introduced over about 1 second and voltage must fall over about the same period or an IMA code will result. Voltage must be around ~140V to start with for it to work.

Your best bet is a standard good battery and switch board feeding a standard MCM for testing IMHO. That's what I will be using.

The stock MDM can probably operate at a much lower voltage it's a fairly dumb drive module, but you have to drive it with all the right signals from the MCM which includes the IGBT drive signals. Very difficult IMHO.

 

[Hugh-Falls]

Assembly drawing

Peter: In a previous post you asked about "assembly strategy". Did you mean a drawing showing how the parts are assembled like the drawing done by Chris(MA)? I never made such a drawing and have not felt the need to do so. The stacked casings and bearing plates can only be assembled face to face so there is no need to show that. The rotors and bearings are assembled on a shaft like a shishkabob so what is important are the relationships between the parts which is just arithmatic. For instance, if you know the "gap" dimension, then you must subtract the thickness of the rotor adapter you made to arrive at the thickness of a spacer to maintan the proper relationship ("gap") between rotors in the stack. The same applies to bearings, flanges, and bearing plates. That is where you need to know also the "offset" and "protrusion" dimensions. It is a matter of having it all add up correctly in the lengthwise direction.

Seeing the drawings of the parts will help and I will post them.

 

[Hugh-Falls]

Addendum No. 2

After adding the additional empty IMA motor case to the end of the stack it looked like a good time to check out the motor shaft again due to the added length. The plan was to design for motor rpm below 3000. Since there is no plan to dynamic balance the rotors after assembly some assumptions were made about the possible resulting imbalance. The original 1-3/16" Dia. shaft looks inadequate now and a new 1-1/2" Dia. shaft (sized to take in account added length and added imbalances due to need to machine the rotors now) will be made. The primary bearing at clutch end was originally sized 1-1/2" I.D. so no bearings need be changed since the shaft diameter at the other end can be reduced to fit the existing 1-3/16 I.D. The four rotor adapter bushings will be changed to suit the new shaft.

One remaining problem is how to increase the rotor opening to allow the new shaft to slip through. Dissembling the magnets from the rotors is to be avoided and mounting the intact rotors in a lathe chuck would probably magnetize the machine so some other way will have to be devised to enlarge the rotor openings.

Assembling the stacked motor might be able to be done with a simple pneumatic or hydraulic (garden hose water pressure) cushioning device if the shaft were vertical rather than the guided screw type fixture used by Honda with the shaft horizontal.

 

[retepsnikrep]

I was thinking of motor rpm same as now ~6200 rpm max. The rotors are balanced individually at the factory as can be seen from the machined holes in them of different depths. In the same orientation they should all still be balanced in a stack but i was thinking of assembling the stacked rotors and covering them in cling film to stop swarf getting stuck on the magnets then taking it to my local engine shop to have the whole assembly checked. The inter rotor adapters and end shafts are the thing that will throw the whole assembly out I assume.

 

[Eli]

Will the IMA motor rev to 6,000RPM by itself? I guess I've never tried. It runs out of significant steam by 4,000RPM it seems..

 

[retepsnikrep]

Will the IMA motor rev to 6,000RPM by itself? I guess I've never tried. It runs out of significant steam by 4,000RPM it seems..

Yes asfaik with manual IMA control I can make it rev to over 4000 rpms and that's with it pumping the stopped engine I see no reason for the electronics to have a different rev limit to the IC. I agree torque will be low above 3000 rpm but to make the gearing useful it will need to rev a bit. Of course we may just be able to engage 3rd gear from rest and forget it.

 

[Hugh-Falls]

Thanks, Chris(MA)

Thanks to the excellent drawing by Chris(MA) here on the Sandwiched Ima Motor thread and on his Electric Bug Build blog, his cross section view of the IMA rotor gives us an indication of its internal construction. Removing the entire crankshaft mounting structure may not be any more difficult than attempting to enlarge the opening in it to accommodate the larger diameter 1-1/2" shaft. Also, this rotor through hole modification would give better access to the the rotor/shaft adapter bushing clamping screws.

 

[Hugh-Falls]

Addendum No. 3

The rotor adapter Taper Loc bushings on the new 1-1/2 Dia. shaft appear to be less stable than on the previous 1-3/16 Dia. shaft. Also, the angular accuracy does not appear to be within satisfactory limits. The Taper Loc setup generally appears to be a bit flimsy so I am abandoning it in favor of a more robust (heavier) Split Taper arrangement and as a result, a new set of rotor adapters will be machined up.

The Split Taper arrangement uses keys to maintain angular alignment which should help with the rotor synchronization and also add increased rigidity (at a weight penalty of almost 3X over the Taper Loc setup). Another advantage of the Split Taper is that by using the P2 longer bushing instead of the normally specified P1, it is possible to eliminate the spacer bushings between rotors. Any gunsmith will attest to the advantages of using fewer parts.

Correcting and erasing actual steel parts in a design can be a bit costly but it is better to make the corrections now rather than later. Fortunately, I don't have to convince anyone other than myself. I may end up with parts for an additional less satisfactory stacked IMA motor before I am done.

 

[Hugh-Falls]

Addendum No. 3 continued

The use of the P1 rotor adapter assumes the modification of the rotor by removal of the crankshaft mounting structure which results in a through hole at the center of the rotor. How the removal is done will be described in another post.

The P1 adapter is made from a #50 "P1" type Split Taper sprocket with 26 or more teeth (50P26). The sprocket is mounted on a short section of shaft with its "P2" bushing and key and the sprocket hub is machined to be a tight slip fit into the flywheel end of the rotor and holes are drilled for what were the six flywheel mounting screws plus a very accurately positioned hole for the flywheel locating pin. The adapter is keyed directly to the shaft (the 1-1/2 Dia. tapered "P" bushing is slotted for the key). The angular relationship between the rotors and the sensor disk must be maintained accurately so the relationships between shaft key and the locating pin will be held accurately on the four rotor adapters and the flywheel sensor disk. A drill jig is made to ensure accurate duplication of all the related parts. The drill jig is made from a fifth "P1" type sprocket. Its holes will be a mirror image of the adapter sprockets because its face will be clamped up against the face of the adapter as they are both mounted and keyed to a short length of shafting for the drilling. To simplify things the locating pin hole and the center line of the key will both be on the axis of symmetry.

The tapered keyed I.D. of the adapter (sprocket) is pulled over the Split Tapered bushing by the three clamping screws squeezing the bushing onto the shaft resulting in an accurate aligned and concentric no clearance fit.

The three hex head clamping cap screws used with the Split Taper bushings will be replaced with socket head cap screws because there is no room for hex heads within the rotor bore. The "P2" bushings are used because they are longer than the "P1" sprockets and will extend out beyond the adapter. The over all length of three of the bushings will be reduced to the IMA motor case width dimension to maintain proper rotor spacing (these bushings are mounted on the shaft tight against one and other) and the fourth bushing length will be machined to suit the space for the clutch end bearing.

By using standard economical power transmission components it is possible to achieve a high degree of precision in the assembled IMA stack without straining the capabilities of a good local machine shop. These parts are readily available from industrial surplus at greatly reduced prices if an effort is made to find them.

 

[Eli]

I did a little test a bit ago while I was doing an in-driveway(in neutral) discharge test, and couldn't get the IMA motor to rev the engine over ~4100RPM, even held at a full 10kW of assist.

But maybe it would go higher without the engine in the way?

 

[Mike Dabrowski 2000]

Eli,
May be able to crank faster with the spark plugs out?

 

[retepsnikrep]

Eli,
May be able to crank faster with the spark plugs out?

I agree I bet the pumping/friction losses are the reason it doesn't rev very high as is. Freed from the IC engine I bet it is very sweet I'm going back to see my engineer this week for an update on assembly plans.

 

[Hugh-Falls]

Other factors?

Probably the Revs/Volt constant may have something to do with the maximum RPM available or more likely in this case, the maximum RPM is limited by the upper limits (switching frequency) of the control electronics. Maybe both.

 

[Hugh-Falls]

Keeper

Magnet keepers are soft iron or steel bars placed across the two poles of a magnet to complete the magnetic field circuit. When the field is contained by the keeper, ADDITIONAL METAL IS NOT ATTRACTED TO THE MAGNET. The eighteen stator iron cores of the IMA motor function as keepers for the twelve bar magnets on the periphery of the rotor at rest. When the rotor is removed from inside the stator, the fields of the bar magnets are open and will attract magnetic material and will gradually dissipate. The rotor magnets need keepers.

If we are to machine the rotor successfully, we need a keeper that can be 1.) clamped tightly onto the rotor and is 2.) easily removable. The keeper will have to be 3.) a tight fit on the shimmed rotor and 4.) be capable of being opened up so that it may be slipped off the rotor easily without damaging the magnet retaining strap using Mike's method. The keeper will need to be 5.) symmetrical so that it does not unbalance the rotor during lathe work.

 

[retepsnikrep]

I think the length of time without a keeper is probably also a factor, I agree if we leave the rotors out of the stators for days/weeks then yes we need something. A few hours or so whilst I machine the rotors probably not a problem. IMHO of course. I'm not going to bother to make anything as I'll keep the time to a minimum. They must have been seperated at the factory for quite some time during manufacture.

I'm going to cover my rotor with cling film during any time out of the stator. so when the film is removed later it will take any swarf etc that has stuck to the magnets with it. the weight of the film is so miniscule as to not be worth worrying about.

 

[morpheus]

Any updates on this?

What are you using to control the motors? Stock IMA system or an aftermarket controller?

 

[retepsnikrep]

Motors are still at the being measured and assembled stage. i will be using stock parts albeit modified.

 

[Hugh-Falls]

Quad stack progress report

The new, more rigid motor shaft and rotor adapters are designed and are being machined. Also, the rotors which need to be machined will be prepared when a satisfactory magnetic field keeper is designed and built.

The initial running tests on the stacked IMA motor assembly will be done using a standard AC commercial variable frequency drive running off of the grid. Also, this same drive can be used in the car as well simply by connecting the battery pack directly to the DC bus, bypassing the AC front end of the controller.

Using the Insight motor controllers, etc. would be ideal but that will have to wait until the stack proves to be a satisfactory motor in the car.

 

[Mike Dabrowski 2000]

I use a 6" wide roll of magnetic steel shim stock. Just wrap the rotor magnets,with the whole roll of shimstock and stop stuff from sticking to it, and it acts like a keeper. Easy to remove by simply unwinding it a layer at a time, as well as acting as a guide when one wants to put the rotor back in the coils.
~ $11 a roll, reusable
McMaster-Carr

 

[jime]

Fascinating thread. Just found it.