Acela 21 (Avelia Liberty) development, testing and deployment

Amtrak Unlimited Discussion Forum

Help Support Amtrak Unlimited Discussion Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I wonder if some of this has to do with the goal of 9in unbalanced vs the 7in the old acela was allowed.
Apparently it has more to do with behavior on straight track, on track that carries mixed traffic (with wheels with different profiles) - things like hunting etc.. The issue of rail and flange profile has been known as an issue since the days of Metroliner, but until they tried to operate regularly at above 150mph they could somehow keep things sufficiently under control and carry on. Still Acelas at 150 mph has a ride quality that no one would accept in Europe. They'd send out track equipment to fix the track, and yet Acelas operate day in and day out bouncing all over the place. This problem has to be solved if we expect to operate trains at 150+mph on mixed traffic trackage. Better to figure it out now rather than after something bad happens.

I am guessing to some extent reading between the lines, so I could be wrong. Maybe @George Harris has more informed insights on this matter, since he has been a practitioner, while mine is just bookish knowledge.
 
There are many things that could be going into the ride quality issues. First some of the non-track non-wheel ones: There is a comfort limit on rates of change in spirals. There are also practical construction maximums to rate of change, but they probably do not apply in high speed tracks. When people talk about raising the superelevation (cant to those that speak British) they tend to blithely ignore this issue. Also, with higher speeds, the rate of change of the rate of change at entry and exit to spirals becomes more significant. To deal with this last issue, the Japanese use a sine wave variation in rate of change of superelevation and radius as well. The French use a modification at the ends of their spirals on the TGV lines to mitigate this, but I don't know their details. Remember, the Pennsylvania Railroad built and upgraded the New York to DC tracks based on an 80 mph speed limit, and may have, if they had good engineers involved, done it in a way that some increase, to possibly 100 mph would be tolerable. At the time the Shinkansen lines were first being built, the "revealed wisdom" in the railroad world was that 125 mph (200km/hr) was essentially the edge of the planet so far as speed on rails was concerned, so it is highly unlikely that anything done by the PRR would even think anything this fast or faster need be considered. If you choose to increase spiral length, as you would need to for comfort at high speeds, that means the offset between the straight line approaches and the arc of the curve will need to be increased, thus modifying the entire curve.

Now as to wheel-rail: The current rail head shape has been derived over the last 150 years by analysis of wear patterns and ride issues. The same is true for the wheels as well. Rails, rail support spring rate, wheel tread shapes, bogie characteristics, spring rates, etc. are a system. You cannot play with just one and not affect the other. This also means that something that works well in some systems will not work well in others. I suspect that has a lot to do with European vehicles on American track issues.

There has come to be a near world wide consensus on the best crown radius for the rail resulting from unrelated independent analysis and observations. The current AREMA rail sections all now have 8 inch crown radii, except the 119RE, which is no longer used in quantity. The 60EN, formerly called the UIC 60, which is essentially the standard European rail section, now has a 200 mm crown radius, which sounds a lot like 8 inches when you consider an inch is 25.4 mm. There are multiple transition radii between this crown radius and the side of the rail head. In normal track the rail is not mounted vertically. It is inclined toward the center of the track. The proper inclination is still somewhat unsettled. The usual in the US and quite a few other places is 1:40. A weight of 60 kilograms per meter seems to have become somewhat of a magic number outside North America. There is the European EN60, a Japanese 60 kg/m, a Chinese 60 kg/m, an Australian 60 kg/m, and may be others as well. The web, base, head, and various connecting radii between these parts of the section are all different. Most of these others had different and larger crown radii the last time I looked, but that has been several years ago. By the way, the conversion between pounds per yard and kilograms per meter is almost exactly 2 to 1.

I know far less about wheels and above, so I will say little about what works here other than a few comments about wheel tread shape. Generally, the thought has been to use a so called "worn wheel" profile. That is, use a wheel profile that makes for a good match to the rail head shape so as to minimize rate of wear. This means that there are multiple radii between wheel tread and side of wheel flange. These radii and a use of a taper in the wheel tread itself results in better centering of the vehicle on straight track and improve tracking on curves. The usual cross slope is 1:20, but other values are used by some. What has been used in the past is a single radius between transition to flange and the outside of the wheel, in other words a cylindrical wheel profile. This has proven to be a bad idea for several reasons. As one example, the first set of WMATA cars had both cylindrical wheels and bogies that were stiff against rotation, the result being wear on the side of the rails in curves being approximately twice that which was calculated in the design. There was probably excessive wear on the wheel flanges as well, but I have no knowledge of that.

I will come back in a couple hours to check for typos and nonsensical statements and grammar.
 
If the problem is ride quality at higher speed, what is the "higher speed" where the ride quality is a problem? 125 mph? 135 mph? 150 mph? or 160 mph? If it's 150 or 160, the stretches of tracks where a train can run that fast are a pretty small percentage of the total length of the NEC, not to mention the stretch between New Haven and New York, where they poke along at less the 70 mph. I wonder if they couldn't give preliminary approval approve the trainsets for service at, say, 135 mph, get them in service to replace the current Acelas, which seem to be falling apart before our eyes, and then keep working on the new sets f or eventual final approval for their designed top speeds.
 
Given that the Acela currently operates at 150mph in numerous sections of the NEC (both north and south of NYC) I'm shocked that certain ride quality issues cant be resolved by looking at the existing acela given they experienced the same euro to US conversion as this generation of trainsets. Sure its slightly apples to oranges but the known/pre existing issues should have been well documented. Even when taking covid into account, the delays here are massive

Apparently it has more to do with behavior on straight track, on track that carries mixed traffic (with wheels with different profiles) - things like hunting etc.. The issue of rail and flange profile has been known as an issue since the days of Metroliner, but until they tried to operate regularly at above 150mph they could somehow keep things sufficiently under control and carry on. Still Acelas at 150 mph has a ride quality that no one would accept in Europe. They'd send out track equipment to fix the track, and yet Acelas operate day in and day out bouncing all over the place. This problem has to be solved if we expect to operate trains at 150+mph on mixed traffic trackage. Better to figure it out now rather than after something bad happens.

I am guessing to some extent reading between the lines, so I could be wrong. Maybe @George Harris has more informed insights on this matter, since he has been a practitioner, while mine is just bookish knowledge.
Given how much of the NEC seems to have newer concrete rail ties in the fastest of sections (NB to Trenton), wouldn't logic say that the flange profile issues would have been resolved at some point over the last 50 years? Or can I read this post as saying this is the same rail that PRR laid down so many years ago with just some new concrete ties on it.

I always thought that the sections of the NEC with new concrete ties, when being repaired or upgraded, were largely now being built to meet standards similar to other high speed lines (albeit being built on existing ROW that sometimes is less than ideal or aging overhead and dated catenary systems), but now I am assuming that isnt true at all.

Not to get too off topic, but are any of these issues being noticed on Brightline's Orlando segment or not really? Sure they arent going 150mph on that, only 125 or 110, but seems like Brightline might inherit some of the same issues Amtrak has that comes with a mixed use railroad and using older track/ROW. Just curious if this is more of an Amtrak issue or just a US railroading issue everywhere. Or perhaps given the amount of construction required, Brightline properly replaced the potential causes of ride quality issues while Amtrak has not.
 
Last edited by a moderator:
Not to get too off topic, but are any of these issues being noticed on Brightline's Orlando segment or not really? Sure they arent going 150mph on that, only 125 or 110, but seems like Brightline might inherit some of the same issues Amtrak has that comes with a mixed use railroad and using older track/ROW. Just curious if this is more of an Amtrak issue or just a US railroading issue everywhere. Or perhaps given the amount of construction required, Brightline properly replaced the potential causes of ride quality issues while Amtrak has not.
well i would think first the brightline train sets have already been tested and approved for use, while the avelia is a brand new train, even the siemens cars with all there issues where first used on brightline as well without many issues. the segment between Orlando and cocoa is brand new while the rest is upgraded freight tracks. i know even in Mansfield where the acela reaches 150 some of the rail dates back to the 80s not to mention other parts that might have rail from the 70s so if there was any advancement in the type of rail the new avelia requires its definitely not on the nec in any meaningful way.
 
well i would think first the brightline train sets have already been tested and approved for use, while the avelia is a brand new train, even the siemens cars with all there issues where first used on brightline as well without many issues. the segment between Orlando and cocoa is brand new while the rest is upgraded freight tracks. i know even in Mansfield where the acela reaches 150 some of the rail dates back to the 80s not to mention other parts that might have rail from the 70s so if there was any advancement in the type of rail the new avelia requires its definitely not on the nec in any meaningful way.
age doesn't really matter for rail so much as how worn it is. there is 75+ year old rail on 80mph segments on the west coast but because they have light traffic the rail is still in good condition.
Not to get too off topic, but are any of these issues being noticed on Brightline's Orlando segment or not really? Sure they arent going 150mph on that, only 125 or 110, but seems like Brightline might inherit some of the same issues Amtrak has that comes with a mixed use railroad and using older track/ROW. Just curious if this is more of an Amtrak issue or just a US railroading issue everywhere. Or perhaps given the amount of construction required, Brightline properly replaced the potential causes of ride quality issues while Amtrak has not.
Siemens has long figured that out, they likely used the ACS-64 to get a good 130-140mph wheel for mixed traffic lines here in the US
 
Age is of no real significance in rail. Rail life is based on wear and development of internal defects in the steel and is usually defined in million gross tons of traffic. Rail is ground regularly to deal with undesirable wear patterns and imperfections in head shape occurring under traffic. Rail is also regularly tested for internal defects in the steel. Rail steel is considerably harder than that used in steel beams. Current rail on the NEC is either 136RE, 141RE, or 140RE, that is weighing those many pounds per yard. I listed 140RE last deliberately as that is considered obsolete. This was originally a Pennsylvania design and many miles were installed in the NEC, so there may be quite a bit still there. "RE" is American Railway Engineering and Maintenance of Way design shape. These are actually the world's best in rail shapes. As rail wears, it is usually "cascaded" to lower speed lower traffic lines. Rail up to 100 years old can be found in minor tracks. Much of this older rail will be in lighter sections, down to as light as 90 lbs/yd. Rail is usually left in main tracks longer than in years past, an in the days of use of jointed rails, end batter and defects around the bolt holes usually defined when rail needed removal from high speed main tracks. Thes rails were usually cut back beyond the old joint bar holes and new holes drilled, and the rail placed in a lower density track. Frequently the process would be repeated once or more additional times.
 
Age is of no real significance in rail. Rail life is based on wear and development of internal defects in the steel and is usually defined in million gross tons of traffic. Rail is ground regularly to deal with undesirable wear patterns and imperfections in head shape occurring under traffic. Rail is also regularly tested for internal defects in the steel. Rail steel is considerably harder than that used in steel beams. Current rail on the NEC is either 136RE, 141RE, or 140RE, that is weighing those many pounds per yard. I listed 140RE last deliberately as that is considered obsolete. This was originally a Pennsylvania design and many miles were installed in the NEC, so there may be quite a bit still there. "RE" is American Railway Engineering and Maintenance of Way design shape. These are actually the world's best in rail shapes. As rail wears, it is usually "cascaded" to lower speed lower traffic lines. Rail up to 100 years old can be found in minor tracks. Much of this older rail will be in lighter sections, down to as light as 90 lbs/yd. Rail is usually left in main tracks longer than in years past, an in the days of use of jointed rails, end batter and defects around the bolt holes usually defined when rail needed removal from high speed main tracks. Thes rails were usually cut back beyond the old joint bar holes and new holes drilled, and the rail placed in a lower density track. Frequently the process would be repeated once or more additional times.
Apparently regular precision grinding of rail heads also prolongs the life of rails and reduces the chances of fatigue failures. I don;t know the exact Physics of it, but it has to do something with micro fatigue cracks of the railhead being prevented from growing as a result of being ground out, or something like that.

These days often you will hear a very high pitched whine while running at speed on some track segments. Those are the ones that have had their heads undergo precision grinding recently.
 
Age is of no real significance in rail. Rail life is based on wear and development of internal defects in the steel and is usually defined in million gross tons of traffic. Rail is ground regularly to deal with undesirable wear patterns and imperfections in head shape occurring under traffic. Rail is also regularly tested for internal defects in the steel. Rail steel is considerably harder than that used in steel beams. Current rail on the NEC is either 136RE, 141RE, or 140RE, that is weighing those many pounds per yard. I listed 140RE last deliberately as that is considered obsolete. This was originally a Pennsylvania design and many miles were installed in the NEC, so there may be quite a bit still there. "RE" is American Railway Engineering and Maintenance of Way design shape. These are actually the world's best in rail shapes. As rail wears, it is usually "cascaded" to lower speed lower traffic lines. Rail up to 100 years old can be found in minor tracks. Much of this older rail will be in lighter sections, down to as light as 90 lbs/yd. Rail is usually left in main tracks longer than in years past, an in the days of use of jointed rails, end batter and defects around the bolt holes usually defined when rail needed removal from high speed main tracks. Thes rails were usually cut back beyond the old joint bar holes and new holes drilled, and the rail placed in a lower density track. Frequently the process would be repeated once or more additional times.
In the late 1980's in Denver we found that what looked like a slightly lower elevation third track on the Joint Line was actually the main line of the Denver, South Park and Pacific narrow-gauge, with rails stamped as being from the 1890's. In a newer use in Denver, alert visitors on the Southeast LRT lines may find Pueblo-rolled rails stamped as having come from steel scrapped from old Mile High Stadium. Properly maintained, they should last longer than the stadium did.
 
Passed by the Race Street yard on #43a few hours ago. One of the trainsets had multiple undercarriage doors open for an unknown reason. No workers present, though.
 
Age is of no real significance in rail. Rail life is based on wear and development of internal defects in the steel and is usually defined in million gross tons of traffic. Rail is ground regularly to deal with undesirable wear patterns and imperfections in head shape occurring under traffic. Rail is also regularly tested for internal defects in the steel. Rail steel is considerably harder than that used in steel beams. Current rail on the NEC is either 136RE, 141RE, or 140RE, that is weighing those many pounds per yard. I listed 140RE last deliberately as that is considered obsolete. This was originally a Pennsylvania design and many miles were installed in the NEC, so there may be quite a bit still there. "RE" is American Railway Engineering and Maintenance of Way design shape. These are actually the world's best in rail shapes. As rail wears, it is usually "cascaded" to lower speed lower traffic lines. Rail up to 100 years old can be found in minor tracks. Much of this older rail will be in lighter sections, down to as light as 90 lbs/yd. Rail is usually left in main tracks longer than in years past, an in the days of use of jointed rails, end batter and defects around the bolt holes usually defined when rail needed removal from high speed main tracks. Thes rails were usually cut back beyond the old joint bar holes and new holes drilled, and the rail placed in a lower density track. Frequently the process would be repeated once or more additional times.
Didn’t the PRR at one time have 155 pound rail? I believe it was the heaviest…
 
Didn’t the PRR at one time have 155 pound rail? I believe it was the heaviest…
Yes, and yes.

The 155 lb section was a Pennsylvania Railroad standard section designed and adopted by them in 1946 as the 155PS. It was a slight revision of an older 152 lb/yd PS section. The 155PS was last produced sometime earlier than 1963. I understand that there was quite a bit in several of the heaviest used Pennsy tracks, including the NEC. So far as I know, all has long since disappeared from the northeast corridor, and probably everywhere else, as well. Apparently it was a step too far so far as track stiffness was concerned.

The 155PS section was HUGE: 8 inches high with a 6 3/4 inch wide base and a 3 inch wide head.

There are heavier rails, but they are designed and used to support rail mounted cranes. So far as i know they have never been used in railroad service.
 
Yes, and yes.

The 155 lb section was a Pennsylvania Railroad standard section designed and adopted by them in 1946 as the 155PS. It was a slight revision of an older 152 lb/yd PS section. The 155PS was last produced sometime earlier than 1963. I understand that there was quite a bit in several of the heaviest used Pennsy tracks, including the NEC. So far as I know, all has long since disappeared from the northeast corridor, and probably everywhere else, as well. Apparently it was a step too far so far as track stiffness was concerned.

The 155PS section was HUGE: 8 inches high with a 6 3/4 inch wide base and a 3 inch wide head.

There are heavier rails, but they are designed and used to support rail mounted cranes. So far as i know they have never been used in railroad service.
Now that’s what I would call “high iron”, 😁back when the PRR called itself, “The standard railroad of the world”.
And their ads touted, “The finest and heaviest railway…”🙂
 
Out of curiosity, if the testing wrapped up tomorrow and got signed off on, how many Liberty train sets have already been delivered and how many are we still waiting to receive? Are we expecting to get the last set within the next 12 months?
 
Back
Top