When doing Beta testing on a repaint a problem has been found with the coupling of this locomotive. The loco will couple with freight cars and other NON GP40-2s but not with GP40-2s from the same pack. This affects coupling nose to nose, nose to tail and tail to tail and results in derailment and thus end of scenario. The solution lies in the .Bin files and this can be found in a downloadable form on the RWA forum: http://railworksamerica.com/forum/viewtopic.php?f=29&t=24512&view=unread#unread All locomotives in the pack and all existing repaints require the modification, presumably repainters will (in future) use the modified .Bin files. I have not posted the actual modification as I have not requested permission to do so however the RWA does not make a charge.
I had already fixed that last year Discussion started with the GP60 and evolved... https://steamcommunity.com/app/24010/discussions/3/3422186614294191177/
Yeah. I see buzz456 took a different approach, keeping the slackless couplers and moving them a bit outside of the box. My way was to just use my standard coupler fix, changing maxdistance and targetdistance, the target coupling point is now outside the box too without changin the engine .bin, only the coupler's. Decide what you like better for yourself. Changing the bin was only necessary for the GP60 as its box is off-centered (and all coordinates are referencing the center of the box (length/2). Here's a diagram I've painted to make things clearer... if the receiving (coupling) point is within the red box, you'll have trouble coupling. The blue parameters are within the coupler blueprint, all others (red and green) in the engine blueprint. Note: "Coupling Pivot" is called FrontPivotX / BackPivotX in the blueprint Link: https://forums.dovetailgames.com/th...-us-coupler-issues-updated.50579/#post-614441 So basically buzz is moving the whole coupler, I'm just stretching it to my standard values (10 cm / 3 inch slack)
So let's have a look at the blueprints and see what is happening to fully understand the issue... Blueprint: DTG\GP40-2Pack01\RailVehicles\Diesel\GP40-2\UP\Engine\GP40-2_UP_Running.xml So here we have the point where the coupler is mounted to the locomotive. We only care for the Z coordinate (forward/backward along the loco axis). Positive is forward from the centre to the front, negative would be backwards. We have the first values of importance here. 8.25, and for the rear coupler -8.25. Nicely centered. These are the inner blue circles in my diagram above. These are the pivots where the couplers are mounted, and center of rotation when driving through curves. Note that the following Receiving Points are not of interest for buckeye couplers. So now we need to know where the collision box is. Let's look further. To be continued... (Actually yes, there is an issue. The collision box is 17.6 m, divided by 2 that's 8.8. The pivot is at 8.25, which is 0.55 m inside of the box. My coupler has a target distance of 0.5, which would then be at 8.25 + 0.5 = 8.75 - still INSIDE so yes the engine has to be modified!) To make things short, change 8.25 to 8.425 (-8.425). This is the culprit. It is the correct value Smokebox would use (8.8 +- 0.375). Change FrontPivotX and BackPivotX from 9 to 9.175 - I'm using these now along with the posted standard coupler fix. Now some might argue the coupler is sticking out too much - to change that you could also shrink the collision box (17.6) and recalculate everything... I'll leave that up to you. As it requires thorough testing (sharp curves)... If you leave the pivots at their defaults (8.25, Front/BackPivot 9) only changing the CollisionLength to 17.25 would also do the trick... a lot of things to change really you would need the exact measurements of that loco. I'll check out TSW2020's GP40-2 for comparing to reconstruct the exact position of the coupler.
OK, I've set up a roadmap for myself. 1. The coupler position is quite accurate in DTG's engine. I will stick to the CouplingPivot value of 8.25. 2. The Front/BackPivotX is where the coupler (front end pivot) will be when coupled (+ TargetDistance in the coupler blueprint) It should correlate with the actual coupler model's dimensions. I'll stick to Smokeboxes' reference of 0.75 metres in length. 3. The collision box length is the variable to change here for not having the buckeye stick out too much. Already done - keep the pivot values as they are and only change <CollisionLength> from 17.6 to 17.25. This will work with both the default and my coupler! To finish what I've started: <FrontCouplingPivot> : where the coupler is attached to the loco - keep it to 0.375 lesser than CollisionLength/2 to have the uncoupled coupler appear at its correct position (it is very close to the loco's body) <FrontPivotX>: where coupler heads are when coupled together (+0.375 outside of the collision box ideally) The coupler blueprint Distance's define the movement under force. I'll upload a complete fix for all engine blueprints and the coupler slack.
So here's the result: Checking the TSW GP40 model My fix - only change is CollisionLength to 17.25. Included coupler fix (Min -0.05, Max 0.05, Tgt 0) for needed slack action to prevent coupler breaking The fix is on the Coupler thread
I have checked buzz456's fix. It works too, but increases the locos' distance between them. Editing the ReceivingPoint matrix is not necessary for couplers of "eMidPoint" type (for 3-link it's relevant) . The FrontPivotX of 9.13 did the trick for him - just moving the coupler away from the loco. RWA fix (compare to my picture above)
Interesting thing is that there's collision box issues with both the GP60 and the GP40, which were done by Reppo. Outstanding models and cabs, equal to TSW standards.
Reppo has done some superb work but someone needs to do more extensive Beta testing to avoid these issues. It takes the shine off when you are limited as to how you can marshall locos due to coupling problems. If someone tries to use a GP40 or GP60 on Falmouth Dock curves then serve them right but coupling to a sister locomotive should be a given
Indeed. You can't be a genius in all disciplines - modelling, texturing and setting up the blueprints. But his models are really top notch. I have to thank you, John, for bringing this coupler issue up again. It has lead me to a new and better way of adjusting and fixing coupler bugs. No trial and error moving pivot points. The steps I follow: 1) Adjust FrontCouplingPivot matrix Z value (15th entry) until the coupler is in his accurate position (uncoupled! - check ingame) 2) Now adjust the CollisionLength. Formula is: CollisionLength = ( FrontCouplingPivot + 0.375 ) * 2 3) FrontPivotX = FrontCouplingPivot + 0.75 (Note it is incorrectly named "X" in the blueprint - of course it is a Z value follwing TS's coordinate system) Now for the coupler blueprint, I'm using MinDistance -0.05, MaxDistance 0.05, TargetDistance 0 (this makes the coupler rest at the FrontPivotX when no force is acting upon it.) Although Smokebox is using higher values, I find that the couplers still stretch out too easily no matter how high Damping and Springvalues are set. So I keep them at 5 cm slack in each direction. Will make further tests on Spring and damping using a modified freight car of 100 000 tons that won't move. That's all!
