Steel or aluminium for bodyshells?

[squizzler]

I understand that most modern passenger coaches or multiple units in the UK are built from aluminium. However Modern Railways has recently discussed the new DB ICE 4 (formerly ICx) and makes reference to its steel body shells.

Intuition suggests that steel is a retrograde step in being a denser material than aluminium and can be prone to corrosion. I note the length of the ICE4 coach is 28m so maybe steel is chosen due to the greater bending forces? Fire resistance? Does the large production run favour designing for steel? Or is it due to designing for production such as factories or plant capable of this order being those equipped for working in steel, or having sunk costs in steel tooling that can be amortised in the ICE 4 order?

So is the ICE 4 an aberration or is steel likely to be favoured for future UK rolling stock?

 

[najaB]

So is the ICE 4 an aberration or is steel likely to be favoured for future UK rolling stock?

Steel is heavier and more prone to corrosion. Both of which argue against its use. On the plus side it is easier to work with - welding aluminium is particularly challenging.

 

[Domh245]

Steel is generally chosen for it's superior strength. Fire resistance wise, they're fairly similar I think. The previous Steel Bodied stock was the Voyagers/Meridians (I think) and given that since then we've seen similar units (in the AT300s) built in Aluminium, I'd be inclined to think that we won't see many more steel bodied vehicles, indeed the in-choice manufacturers (Bombardier Derby, Siemens, Hitachi, CAF, Stadler) all prefer Aluminium I think.

 

[squizzler]

Steel is generally chosen for it's superior strength. Fire resistance wise, they're fairly similar I think. The previous Steel Bodied stock was the Voyagers/Meridians (I think) and given that since then we've seen similar units (in the AT300s) built in Aluminium, I'd be inclined to think that we won't see many more steel bodied vehicles, indeed the in-choice manufacturers (Bombardier Derby, Siemens, Hitachi, CAF, Stadler) all prefer Aluminium I think.

I understand that the AT Hitachi AT300 et. al. stands for "Aluminium Train".

AS you say, Bombardier and Siemens make aluminium trains all the time. It may be that steel is easier to mass produce but even the motor boys have belatedly moved to aluminium for some motorcar models. I would assume it makes more not less sense to master the more sophisticated material for a big production run.

One final advantage for steel is that it should be easier to repair. Even there the IC4 contract apparently includes two spare end cars so DB have apparently catered for the occasional prang...

So I cannot see the logic of going with steel, especially for such a prestigious train for the German industry.

 

[Taunton]

Cars are (very slowly) moving on from steel to aluminium as well, and then to plastic composites for certain parts.

Aluminium is good for :

Lighter (for a given strength)
Avoiding painting
Avoiding corrosion (although not completely)
Avoiding surface rippling long term

Steel is good for :

First cost
Easier manufacturing, both skills and equipment
Much easier repair
Smaller and one-off production runs

The USA has long favoured stainless steel, particularly since Budd developed their Shotwelding process, and the majority of their vehicles since 1945 have used this, both Main Line and Subway, and are wholly or partially unpainted. Not sure why this approach works for them but not anywhere else.

 

[Emblematic]

The ICE4 does use steel, but it's very much a state-of-the-art product using modern techniques, such as laser welding of thin ( varying up to 2mm,) tailored blanks for the body panels. One of the reasons is to avoid the bulk that aluminium extrusions would need to give the needed stiffness on a 28m car length. Longer trains usually sacrifice width to stay in gauge, by avoiding the thickness of the extruded aluminium panels they can keep the interior as spacious as possible.
It's common to think of aluminium as always lighter, but to avoid issues with metal fatigue, load bearing structures in aluminium haveto be substantial box or honeycomb constructions. Siemens in particular have supplied some of the heaviest trains on the network, despite being predominantly aluminium.

 

[D365]

Siemens in particular have supplied some of the heaviest trains on the network, despite being predominantly aluminium.

As far as an explanation for that goes I've always heard that their "Desiro UK" line was over-engineered.

 

[AM9]

... Fire resistance wise, they're fairly similar I think. ...

Aluminium alloys have melting points between 470deg C and 670 deg C, whereas steel (including stainless) melts at around 1500 deg C.
In addition to it's lower melting point, aluminium and its alloys has a very narrow solidus to liquidus range, i.e. it turns to liquid very quickly when its melting point has been reached. Steel behaves much more predictably, going soft and then turning to liquid over a greater temperature range.
On the subject of corrosion resistance, aluminium is very prone to galvanic corrosion owing to its high anodic index. Thus when aluminium is in contact with cast, forged or machined steel alloys which are lower down the anodic index, meaning that the aluminium becomes sacrificial and in the presence of a conducting liquid, as would be the case near the running gear, will corrode quite quickly. Steel on the other hand is rarely used unprotected and modern advanced coatings can reduce corrosion to below those of aluminium.
So really, aluminium only really excels in lower weight per volume.

 

[Shaw S Hunter]

I wonder if there is a "psychological" aspect to this. At the Eschede disaster, which was essentially ICE v concrete bridge, some of the cars were damaged by way of the welded seams in the aluminium body construction more or less "unzipping". ISTR that in the Ladbroke Grove crash the welded aluminium Turbo suffered similar damage. Of course these were extreme events but I believe that in their aftermath there was some discussion about the crash-worthiness of rail vehicles constructed of welded aluminium "planks". Whether such discussions have in fact ever impacted on decisions about train construction I do not know but it's something to ponder.

 

[Emblematic]

I wonder if there is a "psychological" aspect to this. At the Eschede disaster, which was essentially ICE v concrete bridge, some of the cars were damaged by way of the welded seams in the aluminium body construction more or less "unzipping". ISTR that in the Ladbroke Grove crash the welded aluminium Turbo suffered similar damage. Of course these were extreme events but I believe that in their aftermath there was some discussion about the crash-worthiness of rail vehicles constructed of welded aluminium "planks". Whether such discussions have in fact ever impacted on decisions about train construction I do not know but it's something to ponder.

I don't think that accident performance had any bearing on the decision to use steel for ICE4. Both ICE3 and Velaro D high speed sets were introduced after Eschede crash, and both are welded aluminium extrusion bodyshell designs.
Early generations of welded extrusion train bodies are prone to weld unzipping, due to imperfections and brittleness of the weld material. It was a known problem even before the accidents, subject to much research. Modern manufacturing techniques, which include low-temperature welding (such as friction stir welding) and thickening the plate material at the joint, have all but eliminated the weld as a weak point. Any current train will be far more crashworthy than the Networker\ICE1 generation.

 

[HSTEd]

If your fire is big enough and hot enough that your aluminium structure is in serious danger of melting - everyone still aboard the train is already dead.

 

[TwistedMentat]

It wouldn't surprise me if manufacturers switch between the two as manufacturing techniques are developed to improve weak areas.

What I'd be interested to see is if any of them are researching composites for load bearing stuff. With some of the composite tech being developed for aerospace to bring costs down who knows, maybe we'll see carbon fibre EMUs in the not too distant future.

 

[physics34]

I wonder if there is a "psychological" aspect to this. At the Eschede disaster, which was essentially ICE v concrete bridge, some of the cars were damaged by way of the welded seams in the aluminium body construction more or less "unzipping". ISTR that in the Ladbroke Grove crash the welded aluminium Turbo suffered similar damage. Of course these were extreme events but I believe that in their aftermath there was some discussion about the crash-worthiness of rail vehicles constructed of welded aluminium "planks". Whether such discussions have in fact ever impacted on decisions about train construction I do not know but it's something to ponder.

yeh the report into the Ladbroke Grove accident criticised the construction of the Turbos. ....although the accident was extreme the destruction to that turbo carriage was frightening.

Weve replaced heavy underframe overriding risks with a different risk.

 

[John Webb]

Aluminium alloys have melting points between 470deg C and 670 deg C, whereas steel (including stainless) melts at around 1500 deg C.
In addition to it's lower melting point, aluminium and its alloys has a very narrow solidus to liquidus range, i.e. it turns to liquid very quickly when its melting point has been reached. Steel behaves much more predictably, going soft and then turning to liquid over a greater temperature range........

At 500 degC steel will be soft enough to collapse under it's own weight; one reason steel-framed buildings need to have the steelwork protected against fire. But as HSTEd says regarding aluminium, if the coach is that hot anyone unfortunate to be left on the coach will not have survived.....

John Webb

 

[AM9]

It wouldn't surprise me if manufacturers switch between the two as manufacturing techniques are developed to improve weak areas.

What I'd be interested to see is if any of them are researching composites for load bearing stuff. With some of the composite tech being developed for aerospace to bring costs down who knows, maybe we'll see carbon fibre EMUs in the not too distant future.

Maybe for small parts, but carbon fibre/kevlar in quantities required for structural bodywork is probably currently far too expensive for something where there isn't that much of a weight saving, whereas their use in aircraft construction can create large through-life running-cost savings despite a higher acqusition cost.
Then there are low quantity premium uses, e.g. racing/sports cars, cycles, certain military vehicles etc..

 

[squizzler]

At 500 degC steel will be soft enough to collapse under it's own weight; one reason steel-framed buildings need to have the steelwork protected against fire. But as HSTEd says regarding aluminium, if the coach is that hot anyone unfortunate to be left on the coach will not have survived.....

This is true, however before we move on from fire safety I imagine that structural integrity in extreme temperatures might be relevant in determining whether units that run through long tunnels can be driven out with the passengers moving along to other parts of the train. But what material is used for the Eurostar and the Stadler Giruno (Gotthard route) units?

The ICE4 does use steel, but it's very much a state-of-the-art product using modern techniques, such as laser welding of thin ( varying up to 2mm,) tailored blanks for the body panels. One of the reasons is to avoid the bulk that aluminium extrusions would need to give the needed stiffness on a 28m car length. Longer trains usually sacrifice width to stay in gauge, by avoiding the thickness of the extruded aluminium panels they can keep the interior as spacious as possible.
It's common to think of aluminium as always lighter, but to avoid issues with metal fatigue, load bearing structures in aluminium haveto be substantial box or honeycomb constructions. Siemens in particular have supplied some of the heaviest trains on the network, despite being predominantly aluminium.

I don't think that accident performance had any bearing on the decision to use steel for ICE4. Both ICE3 and Velaro D high speed sets were introduced after Eschede crash, and both are welded aluminium extrusion bodyshell designs.
Early generations of welded extrusion train bodies are prone to weld unzipping, due to imperfections and brittleness of the weld material. It was a known problem even before the accidents, subject to much research. Modern manufacturing techniques, which include low-temperature welding (such as friction stir welding) and thickening the plate material at the joint, have all but eliminated the weld as a weak point. Any current train will be far more crashworthy than the Networker\ICE1 generation.

The use of steel seems more puzzling in light of the above, as aluminium construction has been mastered by the main builders and the production facilities have presumably sunk considerable costs into aluminium manufacture - tools training and all the rest. Remember that the industry would have had a considerable learning curve in moving from steel to alloy. If the steel construction used by ICE 4 radically departs from techniques the industry has used before it implies the advantages conferred by switching to steel are sufficient to outweigh the further investment needed to master this process.

I can accept the advantages Emblematic identifies regarding thinner sections fatigue life etc, however the aviation industry which has always benefitted from the best manufacturing techniques money can buy stuck with aluminium for its long tubular fuselages (superficially similar to the tube of a railway body shell) until composites began to take over. This implies it is probably the optimal metal construction. Finally, whist strong lightweight steel alloys exist, I suspect these are not as readily recyclable as aluminium.

Just in case you think I am against steel I am not, merely playing devil's advocate. I am a bicycle enthusiast and favour chromoly steel (i.e. Reynolds or Columbus) over aluminium all day long.

 

[najaB]

I can accept the advantages Emblematic identifies regarding thinner sections fatigue life etc, however the aviation industry which has always benefitted from the best manufacturing techniques money can buy stuck with aluminium for its long tubular fuselages (superficially similar to the tube of a railway body shell) until composites began to take over. This implies it is probably the optimal metal construction.

In applications where weight is the driving factor, yes. For railway carriages weight is probably the fourth or fifth most important factor.

 

[Taunton]

There was considerable aluminium bodying of British rail vehicles from the 1950s. London Underground moved wholly to this, from a range of builders, while the Modernisation Plan dmus brought about a major build of aluminium units from Derby works, who became very proficient at this, from the pioneer Derby Lightweights to Class 108. Given that all of these had notably long lives, the Derby units were generally the longest-lived of all, the technical challenges do seem to have been cracked at an early stage.

A pioneer R49 Stock Underground car was displayed when new at a major industry exhibition, and it's unpainted exterior had been got up to a mirror finish, I read at the cost of many thousands of fine sandpapers used to buff it up. Apparently this was still apparent on this car for years afterwards.

 

[Emblematic]

This is true, however before we move on from fire safety I imagine that structural integrity in extreme temperatures might be relevant in determining whether units that run through long tunnels can be driven out with the passengers moving along to other parts of the train. But what material is used for the Eurostar and the Stadler Giruno (Gotthard route) units?

I think that any situation severe enough to affect the structural integrity will have already disabled the train, cabling, hoses and electronics will have been destroyed at far lower temperatures. The new Velaros on Eurostar are aluminium, so it's clearly not a determining factor.

The use of steel seems more puzzling in light of the above, as aluminium construction has been mastered by the main builders and the production facilities have presumably sunk considerable costs into aluminium manufacture - tools training and all the rest. Remember that the industry would have had a considerable learning curve in moving from steel to alloy. If the steel construction used by ICE 4 radically departs from techniques the industry has used before it implies the advantages conferred by switching to steel are sufficient to outweigh the further investment needed to master this process.

I can accept the advantages Emblematic identifies regarding thinner sections fatigue life etc, however the aviation industry which has always benefitted from the best manufacturing techniques money can buy stuck with aluminium for its long tubular fuselages (superficially similar to the tube of a railway body shell) until composites began to take over. This implies it is probably the optimal metal construction. Finally, whist strong lightweight steel alloys exist, I suspect these are not as readily recyclable as aluminium.

Just in case you think I am against steel I am not, merely playing devil's advocate. I am a bicycle enthusiast and favour chromoly steel (i.e. Reynolds or Columbus) over aluminium all day long.

If something as straightforward as a bicycle has not yet settled on a single optimum material for framing, why would you expect more complex designs like trains to settle on just one? The cabs of aluminium trains are often made of structural steel with a composite shell, so it's not as if the manufacturers are going to lose their capabilities with these materials.

Also train manufacture, like the aerospace industry, has a large and complex supply chain, so any shortfall in capability or capacity is typically met by bringing in new suppliers or subcontractors. In the case of the ICE4:

• body panel sections are fabricated by a relatively new supplier, Photon AG
• the sections are assembled into bodyshells by Bombardier
• bodyshells are supplied to Siemens for the remainder of the build

 

[Emblematic]

There was considerable aluminium bodying of British rail vehicles from the 1950s. London Underground moved wholly to this, from a range of builders, while the Modernisation Plan dmus brought about a major build of aluminium units from Derby works, who became very proficient at this, from the pioneer Derby Lightweights to Class 108. Given that all of these had notably long lives, the Derby units were generally the longest-lived of all, the technical challenges do seem to have been cracked at an early stage.

A pioneer R49 Stock Underground car was displayed when new at a major industry exhibition, and it's unpainted exterior had been got up to a mirror finish, I read at the cost of many thousands of fine sandpapers used to buff it up. Apparently this was still apparent on this car for years afterwards.

There is, however, a big difference between the use of aluminium sheet over a frame and underframe (which in the units you refer to would be steel, aluminium, or indeed a combination of the two) and the modern, extruded aluminium plank monocoque construction. Indeed, you could say that the interesting choice of the ICE4 design is the use of sheet and frame construction to form the bodyshell, rather than the actual sheet material chosen.

 

[squizzler]

If something as straightforward as a bicycle has not yet settled on a single optimum material for framing, why would you expect more complex designs like trains to settle on just one? The cabs of aluminium trains are often made of structural steel with a composite shell, so it's not as if the manufacturers are going to lose their capabilities with these materials.

I don't know where you get the idea that a bicycle might be straightforward (to design! As well as coping with dynamic and static forces with minimum weight, the designer has to balance stiffness for effective pedalling (like a steam loco underframe the bike frame forms the equivalent of the engine block with the riders legs being analogous to the pistons of the steam loco) with controlled flexing (to reduce the road buzz transmitted to the rider).

There is, however, a big difference between the use of aluminium sheet over a frame and underframe (which in the units you refer to would be steel, aluminium, or indeed a combination of the two) and the modern, extruded aluminium plank monocoque construction. Indeed, you could say that the interesting choice of the ICE4 design is the use of sheet and frame construction to form the bodyshell, rather than the actual sheet material chosen.

Do you have an article or something from which you found out about the ICE 4 in detail because I would be interested to look it up. The steel sheet and frame construction sounds suspiciously like an updated version of the Mk3 coach, of which I have seen photos stripped down in workshops!

 

[Emblematic]

I don't know where you get the idea that a bicycle might be straightforward (to design! As well as coping with dynamic and static forces with minimum weight, the designer has to balance stiffness for effective pedalling (like a steam loco underframe the bike frame forms the equivalent of the engine block with the riders legs being analogous to the pistons of the steam loco) with controlled flexing (to reduce the road buzz transmitted to the rider).

OK, fair point!

Do you have an article or something from which you found out about the ICE 4 in detail because I would be interested to look it up. The steel sheet and frame construction sounds suspiciously like an updated version of the Mk3 coach, of which I have seen photos stripped down in workshops!

I've read a few things in print, none of which I have to hand, however online I've found this article which looks to be from the supplier's in-house magazine - so naturally espousing their own processes!

 

[Taunton]

There is, however, a big difference between the use of aluminium sheet over a frame and underframe (which in the units you refer to would be steel, aluminium, or indeed a combination of the two) and the modern, extruded aluminium plank monocoque construction. Indeed, you could say that the interesting choice of the ICE4 design is the use of sheet and frame construction to form the bodyshell, rather than the actual sheet material chosen.

It must have had a significant impact because I believe with the old DMUs it saved about 6 tons weight, or about 20% overall, as the alloy cars were about 28 tons and the comparable steel cars of the same size about 34 tons. The underframe and bogies were still steel so all the saving was in the body.

Weight saving was particularly to the dmu cars which were just powered by low power bus engines (averaging one engine per car). London Underground was separately looking for high performance (which allows more trains per hour), plus reduced maintenance cost and power consumption. Power usage is important in deep tubes where most of the energy ends up as heat to be absorbed or ventilated out after braking. Their previous steel cars had notable corrosion issues, especially the ones which had been stored unused during wartime.

 

[Emblematic]

It must have had a significant impact because I believe with the old DMUs it saved about 6 tons weight, or about 20% overall, as the alloy cars were about 28 tons and the comparable steel cars of the same size about 34 tons. The underframe and bogies were still steel so all the saving was in the body.

Weight saving was particularly to the dmu cars which were just powered by low power bus engines (averaging one engine per car). London Underground was separately looking for high performance (which allows more trains per hour), plus reduced maintenance cost and power consumption. Power usage is important in deep tubes where most of the energy ends up as heat to be absorbed or ventilated out after braking. Their previous steel cars had notable corrosion issues, especially the ones which had been stored unused during wartime.

Although the subsurface stocks on LU were all-aluminium (from the second batch of R stock onward, also saving around 5 tons per car) the tube stocks were aluminium panels with steel underframes, no doubt because there was no room to accommodate the larger frame members that aluminium would require. The 1986 prototype tube stocks were the first to dispense with the steel underframe and use welded aluminium extrusions for the whole shell.
It's notable that there is little difference in weight between the steel underframe and recent monocoque cars, but the latter are undoubtedly stronger and more crashworthy.

 

[Zamracene749]

Could it be that we are looking too deeply here? At the moment aluminium is around 7 times more expensive than steel- 2100 dollars per ton vs 300. 10 years ago it was only twice as expensive- 2000 dollars per ton vs 1100 for steel.

Reading past posts on this thread, the weight savings using aluminium bodyshells aren't as spectacular as one might expect. So using a raw material that is 7 times cheaper and easier to work with must make a considerable saving per train for the manufacturer? Thoughts?

 

[Shaw S Hunter]

Thanks to all for their contributions. This is an area far too often overlooked by us "hobbyists" so it's good to learn things about an area in which I can only claim the knowledge that I've acquired from others. Anyone remember articles many years ago by Roger Ford which discussed the use of Computer Aided Design (CAD) to allow virtual structure testing before any metal was cut? Old hat now but I remember being impressed by the concept at the time.

 

[dubscottie]

You can't compare aircraft to trains. Two different structures doing two different jobs.

On the Derby units, the class 108 were withdrawn pretty sharply when major body problems were discovered after an accident in the very early 90's. The 117's were steel and lasted into privatisation.

The life of a 100% aluminium body for heavy rail is still unknown in the UK.

The first 158's have issues with the castings. At some point, like aircraft they will no longer be safe.

 

[HSTEd]

Could it be that we are looking too deeply here? At the moment aluminium is around 7 times more expensive than steel- 2100 dollars per ton vs 300. 10 years ago it was only twice as expensive- 2000 dollars per ton vs 1100 for steel.

Reading past posts on this thread, the weight savings using aluminium bodyshells aren't as spectacular as one might expect. So using a raw material that is 7 times cheaper and easier to work with must make a considerable saving per train for the manufacturer? Thoughts?

A turbostar vehicle masses roughly 40 tonnes and costs something north of £1.5m

Even if it took fifty tonnes of aluminium to make it (with massive milling losses) it would still only cost a hundred thousand dollars.
Which is something like 5% of the cost of the vehice.

In reality losses are far lower because most aluminium removed from components will be recycled into the supply chain.

 

[dubscottie]

A turbostar vehicle masses roughly 40 tonnes and costs something north of £1.5m

Even if it took fifty tonnes of aluminium to make it (with massive milling losses) it would still only cost a hundred thousand dollars.
Which is something like 5% of the cost of the vehice.

In reality losses are far lower because most aluminium removed from components will be recycled into the supply chain.

Did you mean a hundred thousand pounds??

 

[Zamracene749]

A turbostar vehicle masses roughly 40 tonnes and costs something north of £1.5m

Even if it took fifty tonnes of aluminium to make it (with massive milling losses) it would still only cost a hundred thousand dollars.
Which is something like 5% of the cost of the vehice.

In reality losses are far lower because most aluminium removed from components will be recycled into the supply chain.

Absolutely- of course, the actual mass of the bodyshell will be much less than the mass of the vehicle, so it's probably closer to only 2 or 3% of the total cost. But- if you are tendering for a decent order as has been happening lately, that could easily amount to a 5 or 10 million saving per order- something i suspect the company bean counters might not ignore!