Atomic Systems

I'm trying to get my head around the diesel geared power unit. I'm making a diesel-hydraulic DMU right now and the real train has a very noticable change of engine pitch at about 70mph. From 0-70, it runs in a single gear and from 70 upwards it has a sort of 'overdrive' where the engine RPM drops right back down.
I've got a diesel geared power unit in my blueprint with two gears, with the max speed for the first gear set to 70kmh. With the F5 hud I can see the engine changing gear but the RPM don't get towed back down - it just stays at full blast.
Any ideas how I can get the sim blueprint to drop the RPM when second gear engages or is this yet another thing I'm going to have to try to workaround with a tortuous script?
Are there any locos out there right now that do this?
Thanks !

The Classes 142 (AP/Waggonz), 143 (DTG), 150 (Thompson/Oovee), 153 (JT), 156 (oovee), 158 (DTG & S9BL), & 170 (Thompson) all fit into that category. IIRC they had to fudge the physics as hydraulic transmissions are not properely simulated. I think they used modified Diesel Electric simulation. I have AP sound packs on most of mine, not sure now what they may have done for this issue, but the sound of the revs drop correctly. Dave Dewhust from Waggonz could probably give you some help, or some of the others here more familiar with the inner workings of the sim. Some locos, notably the class 14, 35 & 52 also had Hydraulic transmissions as well.

Oh that's a good point - I could just fake it with the audio blueprint for the sound, but the RPM needs to drop noticably on the gauges too.

I don't think we did it, the sound probably acts like it does but I don't recall scripting anything special for it (it was our first engine)
Id probably attempt it nowadays but not had to yet

Dave

Luckily the 142 doesn't have an RPM guage

DaveDewhurst wrote:I don't think we did it, the sound probably acts like it does but I don't recall scripting anything special for it (it was our first engine)
Id probably attempt it nowadays but not had to yet

Dave

Luckily the 142 doesn't have an RPM guage

Ah. That's a good get-out. Dave - I actually just pinged you an email using your contact form so I guess you can ignore that now then

Scripting seems likely to be the only way around this with any degreee of realism.

A particular problem would be that the real units (assuming a Voith T211 or similar transmission) have two very different regimes: a torque converter, where the engine sits pretty much at its max power speed, and the tractive effort drops with speed (linearly would be a good first approximation); then a fluid coupling, which is much 'tighter' and so drags the engine back down to around its max torque speed, and the engine then accelerates with the vehicle speed, following down the torque curve towards max power / speed again.

The fluid coupling could be approximated quite well with a fixed gear ratio, but the torque converter is more like a CVT (with wildly varying efficiency).

Unfortunately I'm not in a position to share the detailed info I have, but I believe it's possible to download some pretty good brochures from the Voith Turbo website.

ttjph wrote:Scripting seems likely to be the only way around this with any degreee of realism.

A particular problem would be that the real units (assuming a Voith T211 or similar transmission) have two very different regimes: a torque converter, where the engine sits pretty much at its max power speed, and the tractive effort drops with speed (linearly would be a good first approximation); then a fluid coupling, which is much 'tighter' and so drags the engine back down to around its max torque speed, and the engine then accelerates with the vehicle speed, following down the torque curve towards max power / speed again.

The fluid coupling could be approximated quite well with a fixed gear ratio, but the torque converter is more like a CVT (with wildly varying efficiency).

Unfortunately I'm not in a position to share the detailed info I have, but I believe it's possible to download some pretty good brochures from the Voith Turbo website.

This is exactly how the engine and transmission work on the DMU I'm building. I've tried the fixed gear ratio in the geared transmission part of the blueprint but it doesn't seem to do anything - the engine can still spin totally indepent of the speed of the train. I can fake all of this behaviour with a script, including the RPM dial in the cab.
I might have to buy one of the DTG trains and see if I can glean any clues from their blueprints.

You could try buying the BR class 101 pack, which has poor relationship between RPM and gears; but then get the Armstrong Powerhouse pack, to see how Richard Armstrong tweaked scripting to make it more viable.

Paul

I actually have a couple of ideas from other lines of communication. If any of those pan out and become even vaguely usable, I'll drop them into this thread in case someone else comes looking in future

DaveDewhurst wrote: Luckily the 142 doesn't have an RPM guage

They don't need one, if the second stage transmission is kicking in correctly the windows don't drop out!

Regards,

Len

I've just been poking about in the .bin files and found a section like this in both the Kuju 166 and the Oovee 156:
The interesting bit for this discussion is the "Direct Drive: eTrue or eFalse" - this suggests that the torque converter / fluid coupling difference has at least been considered in the game code.

There's an opinion around these parts that the hydraulic model in the game is fundamentally broken, but I've not actually seen a detailed argument to support this. I'm currently taking an interest in DMU transmissions at work, and although I haven't done any real analysis of TS DMU performance, the basics aren't looking totally wrong so far.

On a note that may or may not be related: how would I go about adding an extra gear - how would I generate the next 'd:id="76276064"'? And does anyone have any idea how "TractiveEffortVsGearProportion" is supposed to work - is that basically an engine torque curve?

Yes - if you create your own blueprint using the blueprint editor, you can put as many gears as you like in, with different switch-up speeds and torques. None of it works though. If you do a simple two-speed box with a switchover speed about 70km/h (as an automatic), it switches up at 52km/h and down at 55km/h in the game. I've given up - I'm faking it using scripts (and it's working pretty well so far). The key to all this is a control called VirtualRPM - if that's present, the HUD uses it, and you can tie audio and cab gauges to it too. So irrespective of what the actual RPM is, I can read the game's idea of RPM and then spoof the VirtualRPM control accordingly. That's purely for visual and audio output though - but with some nifty use of the tractive effort CSV file you can fake the change-up there too.

Hmm - disappointing.

What about the Class 180 though? That should have one torque converter and two fluid coupling stages - have DTG managed that in the model?

Looking at what I want to do another way, if I used the diesel-mechanical model, does that have an option for automatic shifts or is it manual-only?

ttjph wrote:Hmm - disappointing.

What about the Class 180 though? That should have one torque converter and two fluid coupling stages - have DTG managed that in the model?

Looking at what I want to do another way, if I used the diesel-mechanical model, does that have an option for automatic shifts or is it manual-only?

All the diesel-geared power units can have manual or automatic shifts. You should be good if you want to do it by hand.