Had a Talgo designed locomotive been pulliing the 501, might the operational mistake been lessened? If a Talgo designed locomotive been operated at the excessive speed, could it have remained railbound and given the passngers a very rough ride? Or, could a Talgo designed locomotive left the rail but the derailment would have seen the equipment coupled, derailed, upright and still within the right of way? Either way, fractures, bruises and lacerations preferable to death.
considering the severity of the overspeed, no the results wouldn’t be much different.
The tilting systems are used for passenger comfort. The tipping speed is a lot higher then the for comfort restricted speed. The tilting lessens the lateral acceleration and force when entering a curve. That makes higher curve speeds possible without exceeding the passenger comfort limits.
Locomotives for tilting trains usually don’t have a tilting mechanism (Acela Express) as the physics are not different and the locomotive engineer can stand the higher lateral acceleration.
Talgo says their system allows up to 25% higher curve speeds. That would have been 7.5 mph for Atk 501.
No, the train would have derailed anyway at 79 mph.
Regards, Volker
An EMD F9 would have tipped at 68 MPH in the 8 degree part of the curve. And, of course, at higher speeds. I din’t have a CG height for the derailed locomotive, so I could not calculate the tipping speed for that one.
If the center of gravity had been appropriately lower and possibly also shifted horizontally, the tipping speed could have been raised to 79 MPH.
These would be the “you can’t go faster because you WILL tip” speeds. It does NOT mean that you won’t tip at a slower.
Seems to me the derailment happened in the eased part of the curve. And it seems to me that then derailment wasn’t caused solely by “tipping”, but also by something that might be called “flange climbing”.
Because the riding flange surface is never at 90 degrees to the centrifugal force as the curve happens, there will always be an upwards component of that force. Which will tend to lift the wheel and vehicle. And change to contact point to a different place on the flange (and rail).
It would be possible to have a derailment based on this flange lifting force EVEN IF THE CG WAS AT RAIL HEIGHT. A special case, indeed. But it illustrates the possibilty of flange climbing contributing to the derailment.
Ed
Than what?
If the locomotive had had a 25% increase in tipping speed:
1.25 x 68 = 85 MPH.
Looks like the “new improved” locomotive would not have tipped in the curve.
Ed
The curve speed was restricted to 30 mph. So the speed restriction would have been 1.25 x 30 = 37.5 mph.
Only the carbody gets tilted to relieve the passengers. The wheel-track system doesn’t change. So the tipping speed is unchanged.
Regards, Volker
The curve restriction would have been whatever management says it would be.
If you can have “25% higher curve speeds”, you can go 25% faster. The absolute upper limit of speed through a curve is tipping speed.
If the fastest an F9 can go through a curve is 68 MPH, and the Talgo locomotive can go 25% faster, you get 85 MPH as a tipping point. The Talgo MIGHT have made the turn at 79 MPH. A train pulled by an locomotive with a CG height similar to the F9 would not.
We don’t know what the design of the engine is/would be. We do know that Talgo moves parts of their cars around for the convenience of passengers. If they moved the CG of the locomotive inwards on the turn, it would raise the tipping speed.
Ed
The speed restrictions for passenger trains are set according physical requirements and passenger comfort design criteria(limit of lateral acelleration and force). In our case it is set to 30 mph according to comfort criteria.
If you are going faster the train exceeds the acelleration and force limits. To drive these speeds without exceeding these limits the carbody is tilted. It is a comfort system not a system to stretch the physis in the wheel-track plane.
The center of gravity is moved to the outside by the Talgo (passive/natural) tilting system. It tilts about 3.5°. To compensate this the carbody is suspended low between the wheels (indipendent suspension) with a lower CG than on conventional cars. Here is a broschure of the Talgo 8:
web.talgoamerica.com/images/Amrak/Series_8_Brochure_sized_opt.pdf
The tipping speed doesn’t get higher.
As said above the CG moves to the outside of the curve. Natural or passive tilting means the tilting point has to be above the CG. Getting down the locomotive floor between the wheels seems impractical.
Talgo offers a locomotive:
https://www.talgo.com/en/rolling-stock/locomotives/travca/
It doesn’t tilt, the power car of the RENFE 130 neither.
In the above linked broschure is a picture showing the speed limit sign for passenger and Talgo trains: P-65, T-75. The derailment curve speed limit is posted as P-30 and T-30. That leeds to the conclusion that the tilting system is inactice
Right. 30 MPH curve. “If you are going faster the train exceeds the acelleration and force limits.”. So, at 31 MPH, the train will derail?
I rather suspect the speed restrictions are set a good deal lower than estimated derailment speeds. Don’t you?
Off and on, we have been speculating about how much slower the train would have had to have been going to have “made it”. I have considered the “tipping” speed because it is an absolute. A vehicle CANNOT go through faster than that. It may or may not make it at a slower speed.
When a manufacturer says a locomotive can go through curves 25% faster (if they did say that) that should mean 25% faster at all speeds. Not just at 30 MPH.
If the upper limit for a “standard” locomotive through that curve is 68 MPH (the tipping speed), why would a 25% faster locomotive not be able to take that curve faster?
You could, and you may be doing so, argue the Talgo engine can crowd the tipping speed much closer than a “standard” engine. This may well be true. The lead engine surely looks like it derailed far ahead of the tightest part of the curve. In this case, the Talgo would have gone farther around the curve before derailing.
I speak of tipping speeds because they are relatively easy to calculate. And they concern such dramatic differences. And it is the absolute upper limit of speed on a curve. Derailments CAN happen at slower speeds. As in the case under discussion, I believe.
We are talking here about the locomotive, not the cars. I do not know HOW Talgo designed their locomotive to go “25% faster”. I did speculate that they COULD have used active movement of CG, not that they did. The simpler would be to use passive (design the locomotive with a lower CG), and they perhaps did.
Or perhaps they
Leads to no such thing: it reflects the fact that the curve has a ‘hard’ 30mph speed limit for geometry reasons, as we have established, and anything going around it cannot exceed that speed. The tilt system may in fact be acting to reduce the perceived centrifugal force or discomfort; in fact, I would assume that it does (since it is passive tilt), albeit the effect is comparatively slight.
There is apparently one Talgo “locomotive” that has tilt, described as the Talgo BT (which I presume is a joint development with Bombardier from the name). I haven’t seen technical material on it, so it may just be the ‘driving trailer’ from one of those four-car XXI sets (the one that looks like a power car but has only a single axle under the nose).
As mentioned, there is little point to tilting the locomotive unless the jacks are arranged to move the CG absolutely inward as the roll is applied. Even there, I suspect there is a limit on increased lateral force before flange contact and then flange climb start to make the gain not worth the risk. True HSR remains dramatically curve-limited, both in horizontal and vertical, no matter how good the transition spiraling can be.
It is not as difficult as you may think to lower a passenger locomotive: take the FM Speed Merchants or the LRC as a starting point, and those both used conventional diesel prime movers. The better question is the one you raised: is the gain from much lower CG worth all the packaging and fuel-bunkering compromises?
There may be another elephant in the room, which I expect to see emerge slowly from testing: is the current Siemens high-speed bogie design prone to quick derailing under ‘abuse’ conditions, leading to derailment before even conventional truck des
In motor vehicle suspensions, it has been held that lower unsprung weight allows for quicker reaction time by the suspension. Hence aluminum wheels, for example.
Less taper would lead to more hunting, and some consequent increased flange wear. I’m not seeing any other problem, though.
I certainly think that the full wheel tread profile (including flange) would be of interest here. Plus also the shape of the railhead.
I’m not finding an overhead view of the crash online, at this time. My recollection is that it looked as if the train derailed very early in the curve, appearing as if it just kept going straight. If my recollection is true, it seems very odd that it did that. Seems way too early. Considering that the curve was almost surely eased. If it was, then it had not hit the tightest part at 8 degrees.
I really wouldn’t be surprised to find the derailment itself was caused by a problem in the truck/wheels.
That does NOT mean I think they would have made it–just derailed later in the turn. Perhaps with more damage/deaths.
Ed
I said the speed limit was set according to passenger comfort requirements. In Germany this allowed lateral acelleration is 0.85 m/sec2.
If the acceleration gets higher the passengers feel less comfortable but the train doesn’t derail.
Did I something different?
Link to Talgo website: https://www.talgo.com/en/rolling-stock/technological-principles/
#2, at the end: For the passenger this equates to a much more pleasant ride, and for the operator it means an increased train speed, since Talgo trains can pass through curves at a 25% higher speed than equivalent vehicles from other manufacturers.
In this case 30 mph is already the fasted allowed (for above reasons) speed. Talgo says up to 25%.
Because the tilting takes place in the carbody only. The result is as if there were an additional super-elevation. But at the wheel-track plane there is no additional super-elevation. The tilting doesn’t change the physics at the wheel-track plane.
My guess is that it left the rails by where the investigators are standing.
Overmod wrote the following post 2 hours ago:
VOLKER LANDWEHR
The derailment curve speed limit is posted as P-30 and T-30. That leads to the conclusion that the tilting system is inactive at 30 mph.
Leads to no such thing: it reflects the fact that the curve has a ‘hard’ 30mph speed limit for geometry reasons, as we have established, and anything going around it cannot exceed that speed. Quote end
That is nonsense. The speed limits in curves are there to protect passengers and loading. This limiting lateral acceleration is much lower than that for tipping rolling stock. As someone, I think 7j43k, calculated an F9 would have tipped at 68 mph.
The Talgo tilt up to 3.5° I think. I have looked at the Bombardier data sheet of the Talgo XXI power car and found no mention of any tilting.
[quote user=“Overmod”]
It is not as difficult as you may think to lower a passenger locomotive: take the FM Speed Merchants or the LRC as a starting point, and those both used conventional diesel prime movers.
I think you mistook what I said. The speed restriction at the point of derailment is dictated by track geometry: it is nominally fixed by rule at 30mph regardless of the type of train, perceived passenger comfort, or whatever. It has been mentioned that this is in part due to absence of superelevation, which is mandated by the few heavy military trains that still traverse the line. It may be important that the remainder of route past the reverse curves is also restricted by rule to 30mph, but in any case the rule, not some conception of acceptable discomfort, is what limits Talgo to the same speed as any other passenger consist at that point on the railroad.
I am well aware of the difference between low functional CG and low roofline or reduced frontal area. I do agree that either prototype I named could have had its CG reduced still further, probably at the cost of some additional sprung weight, were that desirable; perhaps a better example was the RP-210 with the Mekydro ‘powered truck’ which consciously put the components as low as possible.
At the risk of “drifting a topic”, I wonder at the CG height for RDC’s. The two engines were below floor level. And there wasn’t much above.
Ed
Might be. Every curve speed restriction is dictated by track geometry. There is a speed at which the rolling stock will just stay on the track or just tip and there is the speed limit dictated by lateral acceleration limits to protect the load, passengers or freight. The latter is lower. And that is not different at that curve.
Even on a curve without super-elevation a tilt train can go faster than a conventional train. For me the same speed limit for conventional and tilting trains indicate that the tilting isn’t activated at 30 mph. The following publiction shows on page 7, 3rd column at the top, that the tilting system is activated at 70 kph. Sorry it is in German. I didn’t an English counterpart:
www.talgo.de/download/SDNetzel.pdf
I think it is a moot discussion. If you take a current locomotive like the quite low Siemens Charger you can be sure the designers tried to get the CG as low as suitable within the given specification.
Active tilting for DMUs (the whole train) with underfloor engines is an option and there are examples in Europe.
Regard, Volker
difference in speed in a curve in the US is what unbalance elevation is used. The FRA 3" for passenger trains assumes a uniform center of gravity. Freight speeds dictate a lesser 1 3/4" to 2" because the centers of gravity are all over the place combined with spring trucks that are not as stiff under load as the passenger spring trucks are.
See FRA 49CFR213.57 (b) and (c)… The freight railroad ChE’s had varied opinions on the unbalance elevations and how forgiving the curve surfacing conditions could be. ATSF was a lot more conservative than poky old BN.
I visited Talgo’s website, and Talgo appears to say that the tilt is passive, rather than active. So the tilting would always be activated. Perhaps the effect is so little at 30 MPH that the potential increase in speed is too insignificant to make note of.
Ed
Did not help much in the ATSF Redondo Beach incident that soured the railroad on any extensive use of RDCs (up to then I think they had been planning extensive adoption of the technology for a range of services).
It’s not enough to have low CG, you also need enough total vehicle mass to keep the vehicle from tipover. The situation is different from road vehicles, which are limited by lateral adhesion, because sudden flange contact provides a hard fulcrum and even with low CG you can turn over quickly.
I suspect there may have been considerable mass (comparatively speaking) in the radiator structure and coolant high up in the roof, a long lever arm from flange contact.