Performance assessment of IM SD40-2 & Boswer C-636

[llxlocomotives]

This is the thread from the new Atlas motor.  I'm moving it to its own thread because it deserves a saparate discussion.

I'm going to post the results on the IM SD40-2 and the Boswer C-636 in baby steps.

First, the basic physics need to be brought to the fore.
The first chart is representitive and not specific of any our models




The Y axis is tractive effort, draw bar force. The X axis is speed, smph.

The resistance line in this chart can be thought of as the engine only condition on a level surface.

The Propulsion force line represents the maximum speed the train can achieve depending on the load(resistance). This load can be made of train weight, curve sharpness and track grade to name a few. This line is dependent on the system capability (motor & etc.) and the power (average voltage) being supplied to the motor.

The data for 186 engines for draw bar force versus weight is shown in the following figure.




This chart shows the loss in draw bar force that has occured since the year 2000. This date is a little fuzzy. Basically Atlas Kato and P2K engines are Pre 2000 and Atlas China are post 2000.

It is important to keep in mind that these engines are sensitive to external variation. I try to minimize this in the testing. To this end, the same power supply is used. It delivers only a DC signal, no PWM at all. The same 8 foot section of track is used. The track is leveled each week and cleaned daily. My intent is to test only DC engines, but the reality is that even from the factory, most engines have an electronic module of some sort, particularly post 2000. The ones with decoders, like the IM SD40-2 are running with a PWM from the decoder. The decoder sees pure DC imput, but delivers something else. When an electronics module is in place there is additional voltage loss to the motor. I have not attempted to determine what that loss is.

Clearly I need to figure out how to make the charts bigger before preceeding.


I will add to this post and get to how the the two engines Performed in my tests over the next day or so.

[atsfan]

Your charts are not showing up.
What I want to know if what engines and releases have the new motors.
That is really it. Physics are fine. But I want to buy the IM GEVO with the new motor.
Please post a chart of that.
Thanks

[ambluco]

Charts seem fine. Just click them.

[llxlocomotives]

For review purposes, these engines are compared with the overall group and to help the understanding they are also compared with two recent engines that showed to be good performers in the testing.   The purpose is to answer the persistant question of compared to what?

These additional engines are a Kato Amtrak P42 and an Atlas Master Silver Lackawanna Train Master,
These engines are shown in the following photos:

First the SD40-2:


Then the C-636:



Then the P42: (Shown on rollers on the test layout)



And finally the Lackawanna Train Master:


The best picture of the motor in question is from the C-636 engine. It was taken during the truck and fuel tank replacement and the motor did not come out very clear.   



Will start uploading the comparison data tomorrow.

[llxlocomotives]

In this post I will discuss where I'm actually taking data in the operating space. In addition, the forst performance charts will be discussed.

The basic physics chart is repeated here:

The locations where the data is measured are shown on this chart.  You will notice that data is being taken at voltage values of 4, 6, 9 12 and 16,  The max draw bar force is reported at 12 volts.  Data is actually taken at several voltages, but will not be reported here. 

At least 3 data sets are included at each location.  The points in the figures are the average of the three points.  At each point, voltage, velocity and current draw are measured.  At the maximum draw bar force, the velocity is zero. 

The next three charts show the two subject engines in there best light. 

The first happens to be the maximum draw bar force to weight characteristics.  This is actually the same chart as earlier, but a few more points have been added and the four comparison engines are highlighted. 


In these charts, the IM SD40-2 will be light blue, the Bowser C-636 orange, the Atlas Trainmaster is green and the Kato P42 is red.

In this case, the two subject engines do a reasonable job of creating useful force realative to the other post 2000 engines.  Clearly the SD40-2 is the lightest of the four highlighted engines, but it is not the lightest of the recent engines.   One ot the lightest engines on this plot happens to be an Athearn Genesis GP9.   At that weight it pulls the same as the C-636. so its specific power is higher, better.

The C-636 level of draw bar force should pull 17 NMRA weighted 40 foot free rolling cars up a 2.5 percent grade.

The average line slope both pre and post represent the friction coefficient impact of adding weight to an engine.  The polished NS type wheels have a lower coefficient, so the slope is less.  Engines with equal specific power should fall along a line of that slope. With that in mind, the C-636 and the Atlas train master seem to share a common capability. The Trainmaster is heavier and delivers more force. In this case, the SD40-2 seems to be slightly better than the average post engine in capability.  Later I will discuss what happens to these results when I add three weight increments to the engines.  The total weight add was 450 grams or nearly a pound.   But that will be in another post.

Of the four engines highlighted here the P42 shows to be the most robust in delivering force. It does have two motors so that should be expected.

However, none of the four acheive the average of the pre 2000 engines shown in the chart.  If your comparing them with an Atlas Kato, your going to see a significant difference.

The next parameter where the two subject engines seem to do well is the minimum starting velocity.  This is the resulting velocity at the minimum voltage that will support movement.  This is indicative of the engines ability to crawl.  Generally the testing is DC only with no decoder in place.  So this velocity will be higher than can be achieved with a PMW signal. It is indicative of the engines ability to sustain movement at a low power and thus is a favorable attribute. Because of the attention to slow speed running, recent engines have been tailored to achieve low speed low power starts.   



Having said all that, the IM SD40-2 comes with a DCC decoder. This model is no sound version and has the EMU pilot. Because of this the minimum speed result is artifically low compared to most of the other engines in the tests. The other three highlighted engines do not have decoders.

The C-636 and the Train Master both have an excellent low speed chatacteristic in the DC configuations. Because of this they would both be very good candidates for flat switching layouts. The same would have to be said for the SS40-2, with the decoder it crawls at a very slow speed. In this case, the P42 falls on the high side. Still below the average for recent engines, but higher than the other three in the discussion.

Another parameter where the subject engine compared favorablility is the starting current draw. 



In this case all four engines are exceptionally low with the P42 setting the standard for all engines tested to date.

Subsequent posts will discuss areas where the two subject engines did not hold thir own.

[llxlocomotives]

The next set of curves are for the areas where the featured engine do not fair very well.

First the engine only velocity voltage function.



This chart compares the post 2000 engines only.  Missing from this chart is the minimum speed at voltage.  It was show in an earlier figure.  This chart highlights the speed / power short fall of the SD40-2.  The electronic module keep the engine from starting until 6.8 volts.  The problem is that the unit does not have the power to recover from the delayed start.  The C-636 is better, but still is short of a goal speed power relationship.  There are other recent vintage engines with a similar problem.  They are all power short.  It is interesting that the Atlas and Kato engine meet or exceed the goal, because they see both of recent release.

Next The 12 volt velocity from the previous chart is compared with the other two and the total matrix of engines.


This presentation reiterates the speed issues above.  Here the 12 volt speed is the focus.  The tendency for speed to reduce with engine weight is shown for both eras.  The later engines have less reduction with weight, but the average is down several SMPH depending on the engine weight level.  

The SD40-2 speed / power short fall is emphasized.  It is not unique in this problem, but well below the average of both eras.  The C-636 is below but close to the average of the post engines.  The Atlas and Kato both look very competitive.  both have a healthy speed/ power based on this data.  

The SD40-2 and to a lesser degree the C-636 appear to be power short and will have difficulty maintaining a sustained speed at train load.  

Finally, the starting voltage is shown compared the the rest of the pack.

This figure shows the impact of the starting voltage control the on the SD40-2 and other engines from both eras.  Because there is no standard here, the engines tend to not play well with each other in conventional DC.  Yes you can correct this if you strip out the factory electronics or run in DCC only.  It seem like this will limit this engines appeal to DC users.

As has been the case in other parameters, the Kato and Atlas engines start at a competitively low voltage.  The C-636 is at a comfortably low value.  

Different start voltages naturally create a mismatch that have to be tuned in some fashion.

Summary up to this point:

1.- the SD40-2 and C-636 have adequate max torque, but are power short.  This means they will have a difficult time in pulling heavy trains.  

2- problem is common to a number of engines that are from post 2000 releases.  Because of this it is hard to say that one element, like a motor is the primary cause, unless it can be shown to be common to the group.

3- other engines from that same era are much more robust.

4- the data implies a power shift from Pre 2000 to Post 2000 engines. 

5- the C-636 will likely perform much lake the average engine from the post era.  On average they are less powerful than before 2000.

6- the SD40-2 has some definite shortcomings.  It is most likely best suited for low resistance or MU applications.  It would probably do fine in a flat switching layout where it only sees a half dozen cars or so.  

7- I find both the C-636 and the SD40-2 to be quiet, well assembled units on par or better than others from the recent era. I'm not judging the cosmetics, that is best done by others.

The four engines discussed in this discussion were randomly acquired and are one of a kind. There is no indication how they fit a normal distribution in reguard to these results.

In a later reply, I examine these results in what I call a performance characteristic. This attempts to combine the relavent points from the data into an over all comparison. That part of the discussion will be posted sometime tomorrow.

[kcjones]

Holy headache, Batman. I'm still trying to figure out how to program my VCR. And my 5 function Texas Instruments' calculator still works just fine.
I guess I need to learn how to speak Greek.
JL

[theengineshed]

Neat stuff, going to study your results further...

[Brakie]

Aug 3, 2015 18:26:28 GMT -8 kcjones said:

Holy headache, Batman. I'm still trying to figure out how to program my VCR. And my 5 function Texas Instruments' calculator still works just fine.
I guess I need to learn how to speak Greek.
JL

This is why I do my own testing..

All I see is a bunch of squalidly lines I can't make heads or tails of since I'm never studied electrical engineering nor I have found any real need for it in the hobby.

[edwardsutorik]

Larry,

Please describe/illustrate your test rig. How are you getting your numbers?



Ed

[WP 257]

As a professionally licensed (civil) engineer, I get it. Makes perfect sense to me.

Exceedingly well done indeed.

Possibly more personal cost and effort than I'd want to get into to explain the situation, but the numbers seem to very well support the previous conclusions of some of those who have thoroughly run and then upgraded these engines (like Dave aka carrman). The overall trends seem to support what has been said now that we are far into the DCC and sound era--where pure pulling performance of previous generations of motive power has been sacrificed in the name of increased features, sound, and lighting effects.

You've provided real numbers to be able to support what others have been saying, for a hobby in which real comparison data is often sorely lacking.

Again, well done, Larry.

John

[llxlocomotives]

John,
Thanks for you support and kind words.

Ed
I have detailed the test setup and procedure on my blog site.

Briefly, I running on a 2 x 8 foot board. It was sealed and painted to minimize the warpage. I use digital levels to level the track weekly when needed. I clean the track daily. The layout consists of five section of HO flex track. On of which has been designated as the primary test track. All of th engines and this data comes from this track section. The other sections are being used for special tests that are interesting to me.

The data is collected using various techniques.

First I have a DC power supply that provides up to 3 amps of current over a voltage from 0 to 18 volts. This supply has a digital meter that reads output voltage to the tenths and current draw to the hundreths. The supply has been checked using a RRampmeter and agrees very well. I make this check every month or so.

The speed is measured with an Accutrack digital speedometer at the end of the test section. In most cases this is 6 feet from the start. The rate of change in the velocity has been checked and is nominally zero for speeds less than 100 SMPH.

The draw bar force is measured by digital gage and by weight and pulley. They agree very well. I prefer the latter because the kinematics are closer to those of a train. I have found that the 12 volt maximum is close enough. This maximum is when the engine will no longer lift the weight off the floor. I have attempted to measure the point where wheels just begin to spin. It is time consuming and the force difference was less than 5 grams.

I measure the engine weight using a postal scale. I measure the force and weight in grams. The reason is because a gram is .035 oz. most scales for force measure in 0.1 oz. So measuring in grams your reading is 3 times more accurate.  The device might not be, but the reading is.
I hope this helps.

[kcjones]

Larry,
All kidding aside, I see one area where your research will be a great help. With so many mfgs building the same model, performance may be a bigger issue than detail with some modelers. Aka...MTH SD70ace vs Athearn SD70ace. Athearn is more detailed, but MTH will pull the house down. With me... Turn knob right...engine go. Turn knob left...engine stop. And I can't forget my hi tech TYCO "Baby Ruth" sound car
JL

[Donnell Wells]

Given the data that you have posted, what are your ideal characteristics in motor selection, particularly for diesels, i.e., torque, RPMs, voltage, current, etc?

Donnell

[edwardsutorik]

Referring to the very first chart:







I find it odd that the grey line doesn't tend to 0,0--as the blue one does. That implies one could have 50 grams of force with a zero weight engine. Here, I added a red line that does go through 0,0 and still appears, to me, to represent the grey array.


I placed a couple of red X's. They represent two groups of grey engines. Larry, could you speculate on why, for equal weights, the upper group has four times the TE of the lower group.

Note also a similar situation in the blue group. At a weight of 500 grams, there is a range from 55 to 145 grams of TE.



The coefficient of friction between wheels and rail interests me. From old history: I've got a Tenshodo GN S-1 that proudly bears stainless steel tires. The guy's a pulling brute. More recently, a couple of people here have commented that the old Athearn sintered wheels pull better than the current ones.

Interesting.



And thank you, Larry, for sharing your research.



Ed

[llxlocomotives]

Good, the data is getting some attention.

Donnell, I will post an opinion about your question later this evening. I spent the better part of my working career worrying about just that for a different type of engine. My focus has been understanding what things are and not necessarily what they should be.

Ed,
I would say the two biggest reason for the variation at a given weight is actual friction coefficient and motor magnet condition.
When you talking about friction, condition and number of drive wheels matter. So does the weight location. The model.engine designers were more worried about that prior to 1990 or so. After that it seem that it was important to put as much weight on board as possible.

I am one who has the data that shows that stock Athearn and P2K wheels yield more draw bar force than certain polished wheels. I've seen it in several tests where the half wheels were the only change.

Motor magnet condition is another large factor. The main motivation I have in doing this testing is that I rehab HO engines. Thus, I need to know if the problem is electrical or mechanical.  Also for a particular model drive, like a Stewart Kato F unit, how should it perform and what are the acceptable variations.  Generally, any engine can be tuned to be in its acceptable range, except when the motor magnet flux is low.  I have been told, but have yet to varify it, that too many sudden polarity changes will sap the magnet flux.  In the old days, hobby shops would remagnetize the motor.  Today modelers a just buy a new engine.

The reason I went through that is because certain manufactures models tend to be more prone to this problem than others.  Even if it would not constitute as "sick" there is a significant variation that can only be attributed to magnet strength variation.  Lobe or winding count is a factor here.  There are only two poles, north and south.  The thing to do is to increase the number of windings, from 3 to 5 to as many as 7.  All other things being equal, a 5 lobe motor will have 60% of the torque capacity as a 3 lobe.  All things are not equal, so it is not a simple as that, but it does lead to variation in both eras.  

The other point to make is the slope with weight should be made up of two fundamental terms.  The first is the friction coefficient.  That is the average for the system.  The second term is the motor torque capacity either weight.  A pound of extra weight adds a pound times the system friction coefficient that must be produced by the motor.  Eventually the max capacity of the motor will be reached and it no longer will make the required torque.  In that case it will the torque produced will start to fall.

Another thing going on here is that you have apple and tomatoes in one era and oranges and lemons in the other.  

[valenciajim]

Larry--

My dad, brother and son are/were all engineers and they would really have a greater appreciation of this material than would a retired CPA like me. What I get from this is that there are significant compromises being made today that are adversely affecting new products or at least newer versions of previously released locomotives.

Are there specific motors that could be used instead to improve performance?

If this information was more widely available and if there was a Dummy's version, I think your work would do a great service for the consumers in this hobby.

Thanks for posting this.

Jim

[WP 257]

Larry--

Back when I was a student at Penn State, during the very early 1990's, the students in the model railroad club were absolutely obsessive about the Athearn sintered wheelsets. The club layout featured some severe grades - if I recall 4% and more. Some of the guys would routinely replace the Athearn sintered wheelsets after they got some time on them because they knew that at that point the pulling power was compromised.

I always thought they were crazy because my Bowser Santa Fe L-1 Mikado (yes, Santa Fe actually owned 3 of them) would basically pull the house, so I wasn't too worried about changing out wheelsets.

It's interesting that your findings support what those guys knew then--the sintered wheelsets were indeed awesome for traction, until they got worn "smooth" (quotes because I don't know that they ever get all that smooth compared to NS or stainless steel).

I think the naturally somewhat pitted texture of the sintered wheelsets did an awful lot to improve the coefficient of friction, and besides weight and torque, that is one of the key factors in the model world.

John

[llxlocomotives]

John,
Surface roughness matters for friction. The same thing happens when "traction tires" fill up with oil and grit. They become "smoother" and lose some of their grip. Think of real railroads. In many cases the loco will have a load of sand on board that can be "sprayed" on the rails to increase traction for short periods.

The wearing of the sintered wheels is really a polishing of the surface.

Your Bowser Mike had a die cast boiler. That is one of the reasons it did so well. I just tested a Cary F3A shell on an Athearn Blue box drive with sintered wheels. It is off the charts in draw bar force.
Larry

[WP 257]

Larry--

Oh, I'm well aware, but thank you on behalf of others for the explanation...I worked for Bowser during that era if somewhat briefly...the era when how much it could pull mattered a lot more to some folks than how much detail it had on it. Those were simpler days. I well remember the Cary bodies on the blue box drives. We used to run engines for 50 hours just to break them in a little (Anything customers honestly tried to assemble but couldn't finish Mr. English, Sr. or one of the others would finish for them, break in, and then ship back complete and ready for paint. Minimum break in time was 50 hours on the store layout. It was a guarantee that was offered verbally. They took care of anyone who complained either in writing or by phone.)

I just always thought it ridiculous that the Penn State guys would change out wheelsets to gain only a few more cars of pulling power on a severe grade...you know--diminishing returns--lol. They were running mostly diesels, and I had the one steam engine, being a relatively poor student.

John

[Paul Cutler III]

Larry (llxloco),
Not for nothing, but looking up an engineering text will show that sintered metal on N/S has a higher COF than N/S on N/S. That's also true for brass on N/S. On the same loco, sintered or brass wheels will always give higher drawbar numbers because they have "stickier" wheels than N/S wheels. Old locos that have poor plating thickness (like yellowbox Atlas-Kato's or old P2K's) wear down to bare brass pretty commonly and thus will pull better than new engines with 100% of their N/S plating still applied (or ones with solid N/S wheels).

Another issue that affects drawbar is vibration. I've tested hundreds of locos on a spring scale at my club, and engines that don't bounce around as the wheels spin tend to do a lot better pulling than ones that dance around. It's one of the reasons why Katos pull so well vs. Athearn RTR or P2K locos that jerk around like something is out of round (and probably is).

Something else that seriously affects drawbar force is cleanliness of the wheels. Were all wheels cleaned before testing? And were they all cleaned the same way, and just before testing?

You only have 27 data points (engines) for post-2000? And 159 from pre-2000? That kinda skews the numbers just a bit.

One thing I'd like to see added to your charts is amperage. Not that I need to see it, but it would be interesting to know it.

Good work on the data collection.

[llxlocomotives]

Paul,
Based on your first paragraph and the testing I have done, I continue ask my self, "why do I want to use NS wheels?" The only reason I hear is that they are less suseptable to the electrolysis that kills electric conductivity to the engine, particularly in DCC. The performace debit is high and the problem is not linited to the wheel to rail contact point. On both the inside braced Athearn clone type and the outside brace Kato clone type truck designs, the axle shafts show black deposits and it is on the bearings and outside pickup strips, and so on. This problem happens when current is run through metals of different alloy. Again,"why do I want to use NS wheels?" Isn't there a better, less costly way?

That is not really part of this discussion except as you point out it is another variation that is likely in the data, particularly in the pre 2000 engines.

This data is from the last 12 months or so. I have been rehabing engines for a much longer time. My interest finally got the better of me and I decided I needed at least a guide to know what I have done and how it compares with what it could be. The engines generally come in random groups, get what they need and are sold. The exception is specific engines in my collection. Those fall in both era's. The exception to this happens to be three of the engines in this discussion. Thr P42, C-636 and SD40-2 were all acquired primarily for this testing. Two of those will likely stay in my collection. The other one probably will not.

In most cases, the unit will be as received. All get this first test. I have a minimum requirement established. Plus a expectation of a certain model type is developing. Comparing the results to that will be the deciding factor as to what tune up the drive will get. Clearly, if it does not run or the axles are cracked, it gets fixed. If I have the shell off, the bearings and gears get a light lubrication. My actual interest in most cases is to deliver an average model of the type. The closer to stock the better. I go through this to point out that, yes, there is bias in the data. Most of the recent engines are new out of the box. Never had seen any rail time. I test them and unless I find a fault, that is all there is to it. Some are used and require tune up and fixing. Clearly a bias there. The data saved is the best result for each engine.

On my blog, there is a detail listing of the data from these tests. As of this writing, there are 210 engines in the list. At every data point I record current draw. These are tabulated there. Curves are generated for the current results as well. Specific test sequences are summarized in detail on the blog. Again there are charts of almost everything that was recorded.

In a setting such as this, I try to limit the number of charts to those that make the point. The "bleary eye syndrom" can easily over take the audience.

Plese feel free to examine the information posted on my blofg to you hearts content.

[Paul Cutler III]

llxlocos,
Sintered and brass wheels do have certain advantages over N/S wheels. Both are "stickier" and brass conducts electricity better than N/S.

However, both have disadvantages as well. Neither one looks like shiny steel like N/S does. More importantly, sintered and brass wheels have worse dirt problems than N/S does. Sintered wheels are like a sponge for dirt. Their porous surface area retains dirt, and they tend to start throwing off sparks when running. And since arcing causes even more insulating material to form on the wheels, that's not good, either. Brass, OTOH, has an insulating oxidation, while N/S has a conducting oxidation. Meaning that N/S wheels don't need to be cleaned as often as all-brass wheels.

At my club, we have a fleet of club engines from the Athearn BB days. With their old sintered wheels, they wouldn't make it around the layout once without needing a cleaning (headlights start flickering, etc.). After replacement with NWSL N/S wheels, these old engines are far more reliable, and can run for hours during our open houses.

And FWIW, stainless steel wheels are not the answer, either. Something about dissimilar materials having reactions to each other.

[llxlocomotives]

To finish the data charts, I am including some results from several engines where the weight was varied in four increments. Each increment is approximately 150 grams or a total of 450 grams.

These resultas are shown here:



This chart includes three of the featured engines.  The Train master is no longer available, so an Atlas China U36B is shown in its place.  The performance between the two is similar.  Additionally there are four Athearn BB engines in grey and two Athearn Genesis engines in blue. 

Note the difference between the C-636 and the Genesis engines is small, but the SD40-2 looks to be out of gas on torque.  Eventually all engines will show that character with weight.



 

[Great-Northern-Willmar Div]

Aug 5, 2015 10:42:41 GMT -8 llxlocomotives said:

To finish the data charts, I am including some results from several engines where the weight was varied in four increments. Each increment is approximately 150 grams or a total of 450 grams.

These resultas are shown here:



This chart includes three of the featured engines.  The Train master is no longer available, so an Atlas China U36B is shown in its place.  The performance between the two is similar.  Additionally there are four Athearn BB engines in grey and two Athearn Genesis engines in blue. 

Note the difference between the C-636 and the Genesis engines is small, but the SD40-2 looks to be out of gas on torque.  Eventually all engines will show that character with weight.



 

Atlas doesn't make a U36B. They do have a U23B and U30B

[Spikre]

??
Atlas also made the B23-7.
did they also make the B30-7 ?
just wondering ??
Spikre

[llxlocomotives]

You right.  It was a U36C:





The Genesis GP9 is shown here with the weight increments applied running on the test track:






Hope this clarifies what I'm doing.

[Great-Northern-Willmar Div]

Aug 5, 2015 11:33:07 GMT -8 llxlocomotives said:

You right. I ment U36C:

But, the U36C is nearly 20 years old. The electronics and the construction of the drive is different from Atlas circa 2015.

Since we are talking about six axle units SD40-2 and C636 an Atlas SD35 or SDP35 from the last run would be more similar in weight to the Intermountain SD40-2. Can't say what the weight of C636 is as I've never owned the model.

Also, since we are looking at the Intermountain model as being a gutless wonder, how does it compare to the current production runs of Athearn RTR SD40-2, the Kato mid-SD40-2 and BLI SD40-2.

These locomotives drives can change from run to run so comparing the latest and the greatest is a bit more apples to apples and definitive to what is currently being manufactured.

[edwardsutorik]

Aug 5, 2015 11:22:29 GMT -8 Spikre said:

??
Atlas also made the B23-7.
did they also make the B30-7 ?
just wondering ??
Spikre

Yes.



Ed

[edwardsutorik]

Larry,

This latest chart is interesting (well, so are all the others).

One question, just in case: Are all wheels powered on the IM SD40-2. I would assume so, but ya never know until ya know. The AHM U25C had an unpowered center axle, as I recall.

The slope on the IM loco is dramatically undramatic. I assume the motors never stalled on this test. If so, it would seem the wheels on the IM were hugely slippery.

If you're in the mood, I suggest you put the locos on an inclinable track and increase the slope of the truck until they slip. That would be without power--they just sit there until the slope gets too steep, then wooooops! I wonder if the IM would slip "earlier" than the others.



Ed