Message Areas - "Do We Need a Dynamometer Car? Or eng Files Revisted"

Train-Sim.Com Message Areas

"Do We Need a Dynamometer Car? Or eng Files Revisted"

Email this topic to a friend
Printer-friendly version of this topic
Bookmark this topic (Registered users only)
Previous Topic | Next Topic Conferences MSTS Locomotive Design (Protected) Original message
woodman (112 posts)
Jul-25-01, 05:42 PM (EST) "Do We Need a Dynamometer Car? Or eng Files Revisted" While I was testing the F unit eng files I notice some things that I really haven't had time to investigate fully. But, now that I'm receiving reports about the loco's pulling power, I'd like to raise them and get some feedback from this forum. Just two for now.Does locomotive orientation, i.e. one unit facing rear in a power lash up affect performance? If it does, this is a major bug. It also makes B units problematic as well, if they have to have a "front end".Do helpers really add power to a train? I've placed F units at the rear, cut in in front of the caboose, and noticed that they don't spool up in either sound or smoke. Would it be different if they were behind the caboose, pushing. This has implications regarding DPU set ups.So, I'm going to go back to the drawing board. Reinstall a clean installation, install my test units, go over the eng file with a fine toothed comb. Then I'm going to set up a test train of known weight and test them over a time and distance segment. The ability to accelerate a train is a pretty raw measurement but it's a start.What we need is a sort of SimAAR facility with dead smooth test tracks and known grades. But, what I could really use is a dynamometer car or a dyno read out of some kind that would read out number to a file.Mike Woodill
http://www.3dtrains.com Alert Edit | Reply | Reply With Quote | Top
Table of Contents

Subject Author Message Date ID RE: Do We Need a Dynamometer Car? O... HAC Jul-25-01 1 RE: Do We Need a Dynamometer Car? O... woodman Jul-25-01 2 RE: Do We Need a Dynamometer Car? O... stationmaster Jul-25-01 3 RE: Do We Need a Dynamometer Car? O... woodman Jul-25-01 4 RE: Do We Need a Dynamometer Car? O... Jeff Jul-25-01 5 RE: Do We Need a Dynamometer Car? O... CactusBoy Jul-25-01 6 RE: Do We Need a Dynamometer Car? O... Wurger Jul-26-01 7 RE: Do We Need a Dynamometer Car? O... plainsman Jul-26-01 8

Lobby | Topics | Previous Topic | Next Topic
Messages in this topic
HAC (140 posts)
Jul-25-01, 05:59 PM (EST) 1. "RE: Do We Need a Dynamometer Car? Or eng Files Revisted" Hi Mike..
I'll start doing my own tests as well, and get back to you..
I've noticed that DPU does not work in MSTS.. I think I'll run up
a standard "scale test" car tonight and start doing some tests
on the F's, especially in DPU mode..cheers
Harold Remove | Alert Edit | Reply | Reply With Quote | Top
woodman (112 posts)
Jul-25-01, 06:46 PM (EST) 2. "RE: Do We Need a Dynamometer Car? Or eng Files Revisted" Thanks. I'm going to make consists with single units on both ends; a single unit on the rear end and some permutations. I'm also going to test couple of ideas I have. Let's try to use a train of the USxxx7 AT&SF, box cars as standards and keep consistent hp/ton figures.We still need a a Dyno Car.Mike Woodill
http://www.3dtrains.com Remove | Alert Edit | Reply | Reply With Quote | Top
stationmaster (65 posts)
Jul-25-01, 07:24 PM (EST) 3. "RE: Do We Need a Dynamometer Car? Or eng Files Revisted" Model railroaders use a test bench that consists of a straight piece of track on a board that can be tilted to represent various grades. The fancier ones had sensors that could be used to time the locomotive as it climbed the grade and to get accurate speed measurements. Somehow, the details escape me, there were calculations for determining the locomotive's tractive effort and efficiency.Maybe what we need is a standard "Test Route" that could serve the same function.
Remove | Alert Edit | Reply | Reply With Quote | Top
woodman (112 posts)
Jul-25-01, 07:45 PM (EST) 4. "RE: Do We Need a Dynamometer Car? Or eng Files Revisted" I'd be interested in finding out if 4hp/ton would be enough to get a train over Marias. That should be enough to climb the walls of the Kremlin. I'm going to begin testing at 2hp/ton as soon as I have a clean install and a verified eng. file.Mike Woodill
http://www.3dtrains.com Remove | Alert Edit | Reply | Reply With Quote | Top
Jeff (9 posts)
Jul-25-01, 09:18 PM (EST) 5. "RE: Do We Need a Dynamometer Car? Or eng Files Revisted" At first I was very disappointed with Train Sim. It left me thinking here was a game made for the model train crowd rather than a simulation. But after waiting for this game for so long I persisted with it and after a bit of editing of the .eng files was very surprised just how damn close the physics are. Maybe Union Pacific does have a point after all. There are a few problems with the physics side of things though:
· Wheel slip modelling is wrong. No big deal really.
· Air brake needs a rethink. Looks like Kuju didn't have the time to finish it. From what I can tell they modelled air brake as a complete whole instead of a series. ER first, BP second, followed by the individual wagon triple valves. For a quick work around change the max release rate to 0.35 and application rate to 0.20 in the .wag files. As for the wagon braking force I'll have to go through all my air brake books to get the right figures. It all depends on wagon length, weight, brake rigging efficiency, reduction valves blah blah blah.
· There a few very minor quirks. Tractive effort spiking instead of being smooth, Independent brake too weak (why is there wagon braking variables in the .eng files? ).During testing I find a continuous grade on any of the routes, make a train with a known weight and length and see where at what speed it is balancing. Both ascending and descending. The balance speeds where scarily close to real world figures. During one test I changed the Dash 9 figures to what I regularly drive (Who would have thought a 120 tonne narrow gauge 3000 hp loco would have a slightly better continuous tractive effort than Dash 9) hauling a 6000 tonne train. On one route the balance speed was identical to a bank I drive over ever day. Train sim grade: 1 in 77. Real world grade : 1 in 80. Factor in the effects of curvature and real world rail head conditions I don't think you could get any better results. I was very impressed.I've had no problems with remote locomotives powering up on trains. Have tried uncoupling and re-coupling remote locos with no problems.For me the largest problem by far in Train Sim are the grades. Especially Marias Pass. They are all over the place and greatly exaggerated. Real world physics and model railroad track don't mix. Forget about using HP per ton trying to work out what a loco can pull. It doesn't work that way. Using powered axles is ok to a certain extent. Long as the ratio for different locos is correct. If you have the right variables in the .eng files you can be confident the train is running close to real world specs. The problem lays with the track grades. Hope this helps.Jeff
Remove | Alert Edit | Reply | Reply With Quote | Top
CactusBoy (20 posts)
Jul-25-01, 09:30 PM (EST) 6. "RE: Do We Need a Dynamometer Car? Or eng Files Revisted" Noticed the same things myself. Any loco I download, I put through a serious set of paces much like railroads did when testing demonstrator locomotives. Few things I've noticed (other than your points) Include that the engines are slow to start when you have them set up back to back...but elephant style (nose to nose) they start up quickly, spooling up to a good running speed in no time.
Couple other things I noticed, mostly superficial and nothing to do with the engine itself. However, I've noticed, along with many who post here, that pretty much ANY locomotive you have cut in the middle of a train or at the end, never seems to pull its own weight.The other, which may or may not happen to other people (don't know if I'm the only one) happens when I uncouple and recouple to another engine. The second engine (the dummy so to speak) acts like its brakes are locked up and won't move. So, I'm screwed if I end up using anything less than my whole cut of locos to move something. I think that's a MSTS bug though. Remove | Alert Edit | Reply | Reply With Quote | Top
Wurger (69 posts)
Jul-26-01, 11:56 AM (EST) 7. "RE: Do We Need a Dynamometer Car? Or eng Files Revisted" I have not noticed any difference in pulling power being dependent on the direction a loco is pointing in a consist. At least with the modern diesel's. I have not DL'd the F7 yet.I do belive that the physics engine is pretty good on the power side of the locos as far a mirroring real world peformance. I use a test track that I made with the route editor that has level track and a known grade. I then use the US2graincar with the mass @ 119.29t (131.5 us tons) in different quantitys to test each engine. I determine the max continous TE for the loco, plug this into the calculator and determine the max number of cars the engine can pull on a given grade and calculate the speed the train should balance at. I then change the adheasion and numwheels (under power) variables to get the train to balance at its max TE speed (with wheelslip ()). Once it's performing in the ballpark, I set wheelslip (1) and test starting on grade and the balance speed to verify it's working right and then add an extra car to make sure it stalls on the same grade (since it would be exceeding it's max continious TE).Once the basic variables are set you can plug in the correct values into the eng file and each loco balances real close to the predicted speed every time. This leads me to belive the program is doing the math correctly.The process of testing the maxbrakeforce values is even more complicated. If you use the correct (realworld) values for maxbrakeforce, about 13% of GRL for a loaded car, you CANNOT control a train down the 6% percent grades from blacktail to Java. I have data from the GN RR that shows the track profile from the summit to java and it is 2.0 - 2.2% all the way. From what other engineers have told me, you should be able to balance a train on that type of grade with around 10psi of air and DB5 to DB8. I have to use a maxbrakeforce of 45% GRL to get a 50 car grain train down with two dash nines and 26psi set and DB8. This proves that realworld brake performance will be impossible to implement on routes where the grades are incorrect. I do know that helpers on the end of a train do make a difference. I have an activity on Marias where I take a 64 car grain train over the pass EB with 4 dash9's on the point. this train will stall at Java (2.09HPT) unless I pick up three helpers to shove on the end of the train (3.65HPT total). Even then, this train goes up Java at 9-13mph. I can change the directions of the locos in this consist (nose to nose, tail to nose, etc.) and get the same performance.So what's my point you ask?
I don't really have one I guess!
Thanks for listening and keep up the good work Wurger Remove | Alert Edit | Reply | Reply With Quote | Top
plainsman (32 posts)
Jul-26-01, 04:11 PM (EST) 8. "RE: Do We Need a Dynamometer Car? Or eng Files Revisted" Locomotive direction of travel is more of a problem with some engines, and apparently not a problem for others. The Kiha 31 is the most egregious problem in this regard. It is mild mannered in forward travel, but takes off like a rocket sled in reverse. This becomes even more obvious as you increase horsepower to emulate similar types of self powered cars (like the Doodlebug). The GP38 seems to behave just as it should in either direction. I think more time was spent getting the look just right , and less on the engineering and physics and geography.
Bob Remove | Alert Edit | Reply | Reply With Quote | Top

Lock | Archive | Remove

Lobby | Topics | Previous Topic | Next Topic