SDP40: Final Verdict?

Trains Magazine Passenger
21

Please tell me how much height you suppose these tanks to have, since the frame seat was less than 7 feet above the railhead.

I have heard many explanations for the problems that the SDP40F’s had…including the water theory, rail alignment and railhead contour, which by the way, the lines that comprise the Southern Region of CSX experienced most of the problems, and truck hunting, which although hunting has been a word that was used, it ceased when modifications were made to those trucks.

It could have been anything, I suppose, including your water theory. Doubt that, though, and the NTSB and Amtrak reports tend to discount that idea.

Save your scientific reports. The fact is that the locomotives were junk as passenger engines. The fact that Santa Fe did not have problems with them proves little, since they didn’t derail on ATSF track, either, to my knowledge, nor on UP or SP. SCL, L&N and other Eastern Roads had the lion’s share of the difficulties.

And actually, on a 100 car Oil Can, you do feel surge, whether you want to or not. After 31 years as a Locomotive Engineer, I believe I have earned at least the right to know about slack, if nothing else.

I doubt seriously that a computer program without the necessary raw track data is going to solve a mystery that has existed for this many years, but have fun.

Case closed. I have to go to work.

22

+He is using the DESIGN Specs of the SDF40P including weight and were stuff was located he is also going to be using the track standards of the FRA for the speeds plus is going to be using various other things that could of have happened during the usage of the units. His design of the Modeled unit is BASED OFF THE BLUEPRINTS and SPECS so it should be accurate. He is also going to be using the standard maintance standards of the day for trackage 39 foot jointed rail and such. Yes it is not the same as the real life locomotive but it is better than nothing.

23

Let’s have a look at that calculation.

24

In an earlier reply, someone inquired about where the SDP40F-related derailments occurred. From the book I have, several occurred on BN trackage over which the California Zephyr travelled, and others occurred on former L&N trackage.

25

Timez I pulled tankers for a few months and it takes about 40 tons to tip one over. I have the charts someplace and 2000 gallons that weigh 16 grand moving at 79 mph would have that much force. Now that much surge may not seem like alot however you get it going side to side before it settles down you are getting into harmonic vibrations and next thing you know you are on your side. Why do you think tankers take most curves 5-10 mph slower if they are 1/2 full surge. I saw a loaded tanker moving at 60 mph on a wet road get slid into a ditch like some one picked it up and placed it there.

26

Not just tank trucks are a problem. There is one such incident involving an air ride trailer of a famous furniture mover hauling electronics in the Boston area in winter. apparently it was icy that morning and the driver swears he exited the off ramp off 95 at less than the posted speed. Apparently the trailer air ride suspension kept leveling the trailer and the result was a fulll load of computer equipment on its side burning. The resulting claim was several millions of dollars. Of course they fired the driver an owner operator. I also hauled electronics at that time and many of the main frame computer equipment was top heavy at that time and can understand the drivers argument. I never lost a load but came close on a couple of occasions.

Al - in - Stockton

27

You get anyload that is TOP HEAVY or put any heavy oblect such as Water TANKS for a Steam Genarator ABOVE what on a normal unit would be the CENTER OF GRAVITY you are going to have problems. Couple that with bad track from deferred maintance from lack of capital and then a hodgepodge of equipment that might not respond all the same when it came to braking forces and such and LOOK OUT.

28

Hope his physics is better than yours! [:)]

29

He already got it done. He called me tonight the results are in. TURNS OUT IT WAS A COMBINATION OF FACTORS. One was the water tanks like I suspected. Second was the truck design being a hollow Bolster it did not have the mass needed to stop hunting once it started. Last was the trackwork. What would happen according to his simulations 95% of the time is first the engines would hit a low spot and rock side to side getting the water sloshing in the tanks then once that was going the trucks would start to hunt keeping the first thing going and getting worse then throw in a curve and bang rear trucks on the ground. Was worse for them if coupled back to back like I suspected. The other 5% happened do to bad trackwork only given the RR’s Finances at the time I could see that happening.

30

Now this brings up an interesting question. What about water slosh in steam tenders and also oil in the oil burners. Maybe the tenders were sprung differently? Any of our steam operators have any input???

31

What was the peak L/V ratio?

32

Like the vertical and underbelly tanks on the SDP40F’s, locomotive tenders had baffle plates in them to reduce the effects of water and fuel oil sloshing. This in turn reduced the aforementioned forces working upon the tenders.

The 40F’s underbelly tanks had baffles in them from the facotry. It is unclear if the baffles installed in the vertical tanks came from the factory as well, or if the tanks were retrofitted.

33

His data showed that water levels between 75% down to 33% were the most dangerous on them. Fuller than that no real issuse lower and again not enough mass to cause a Problem. His data showed speeds were normal track speeds for the class of tracks and curves they would have been running on. AKA 79 MPH and 90 with ATS with approrite reductions for curves. Now with the tanks Everything I have pulled states they were never baffled WHY I never will know unless they were fearing low water levels shutting downt he heat in the winter. He ran the sims with both baffles and unbaffled and no change. Most derailments took place on his sims between 60-79 mph with the trucks hunting and almost a harmonic ocsilation of the bodies with the water tanks surging back and forth.

34

The belly tanks had baffles; the uppers were retro’d. What does the sim show on CWR on various railhead configurations (ground and unground) on curves better than 2 degree with a 200 to 400 foot runoff and 6 inch superelevation?

35

So, how about some peak L/V values from the simulation?

36

Good Question. The L/V ratios ought to show something…

And while we’re at it, how can cousin get an accurate rendition of what is going on if he doesn’t know the curve radii, superelevation and other such specific factors? And if he does, where is he getting such info? Generalized numbers on jointed rail as edbenton mentioned earlier would be valid only on the Santa Fe’s ATS territory in New Mexico-Colorado-Kansas, where no derailments occurred. Most accidents occurred on CWR, didn’t they? Except one where the track was proven to be out of gauge at an interlocking. There hasn’t been mention of drawbar hp, braking effort, consists, railhead contour, track gauge, cross level and a host of other factors…

I’m impressed. This busy computer programmer was able to write a complex program to simulate all the factors involved in less than 4 days while working his regular job? All the dimensions, factors, loads etc? That is impressive. Maybe he can come on and defend the numbers and explain the program so we can get a better grip on what it is intended to do and how it does it?

Give us the specific data, if possible. That would help us out.

37

I did ask him what he based the train weights on he told me historical consists which means 12-16 cars Approx weights of around 900-1200 tons weight. HE did figure in the Superelevation based on a commercial based program for engineers he did not say anything on the rail grinding or railhead contour. What he was going for was what I asked him were the main causes was it Track the Trucks or the Water tanks or some other factor and what were the fators in the design that COULD HAVE led to the derailments not a definative answer. People I asked my cousin that has no railfan interest at all in fact. He could have told me to SHOVE it he did this because he wanted to not because he had to now you are calling his data into QUESTION. WHAT DO YOU WANT HIS FIRST BORN CHILD TO MAKE YOU HAPPY.

38

No, but some understanding that a complex program such as that is extremely difficult to write in so little time, especially without a full understanding of the forces and variables which existed at that time.

Without allowing for draft and buff forces, such things as L/V ratios as mentioned earlier, and other complex mathematical computations, writing an accurate and reliable program such as described would be virtually impossible in the time frame allowed, unless a game platform were used as its base.

I would just like to know how he did it and what factors were used. I am sure others would be interested as well, since throwing out numbers as an explanation is nice, but without the basis for those numbers, the explanations tend to mean little or nothing. The explanation of the facotors that were and were not considerd would be helpful, and knowing whether another programs basis was used to make an add on program out of it would also be nice to know.

Computers weren’t available in quantity and in the power needed to do the calculations back when the issues were occurring, so it would be beneficial for those of us on the board with computer programming experience, or with access to programmers to get the infor from your man, so we can compare notes…sort of like going to a doctor for a second opinion of a diagnosis. Nothing to get irritated about. We definitely appreciate his effort.

39

It’s hard to call into question what hasn’t been seen.

There is a market for the kind of simulation you describe. Your cousin should market it and publish and present at the ASME/IEEE/ASCE joint rail conference. (here is link to the conf for 2008 which was held the past April. http://www.asmeconferences.org/jrc08/ They haven’t had the call for papers for 2009 yet.) His code is copyright protected - he could make a few bucks.

The numbers that are needed to predict the likelyhood of a derailment are the L/V ratio. That is the ratio of lateral force to vertical force of the wheels on the rail. Those are the numbers that need to be the output of the simulation. After having modeled all that other stuff, getting this as output would be trivial for him.

That’s all I’m asking for. And, no, I don’t want anbody’s first born. I have two teens at home. [:)]

40

I will give him a call and see what he has for those measurements tonight and should have and answer for you.