How about a follow-up about the methods used in the actual Pressing or Extruding process?
The Press needs something to give the part a desired shape. I'm assuming some kind of strong/hard pattern/plate above and below the material to be pressed. What are they made of? How long do they last? How are they made? Is the forming done as a single slow procedure or multiple strikes? I've seen videos of steel railway wheels being forged from rectangular blocks... How is this different?
In Extrusions, how do they get the hollow centers from solid material? How is butted tubing made?
This is a wonderful article. I had no idea that these very heavy presses needed such government support, and it's rather depressing that further progress in this important field of manufacturing will be, in practice, made by the Chinese.
I think most of us think of manufacturing as a boring, low-tech field with doesn't require much R&D, and certainly doesn't require programs to be sponsored by the government. How wrong we are.
When I was pre-teen living near Cleveland, Ohio in the 1960s, Alcoa had a Bring-Your-Family-to-Work day. I got to see the 50,000 Mesta forge press in action. It stood several stories high and also extended several more floors into the ground. When it moved, it appeared like slow motion. Not what I expected. The top die would descend for several seconds. When it impacted the work piece, there would be a shutter...like an earthquake and a low frequency rumble. You could feel this in every building in the area.
Dad graduated from the University of Illinois in Champaign in what was a new field at the time: Metallurgical Engineering. The school also had a large press on campus.
Another thing worth noting about these projects was the tremendous speed at which the presses were designed, fabricated, and put into production. Seems like 3 to 5 years, which included deciding on and obtaining the site, building the ancillary structures, and building and assembling the presses themselves. I don't know if this was the norm back then, but now it seems to take 3 to 5 years just to get the contracts signed for any significant effort.
That was exactly my thought. "Ah, this was back when we actually had the minimum competence and state capability to actually accomplish moderate stretch goals. Those were the days, weren't they?"
Not only that, but they kept the "cost per press" about the same even after dialing it back from 17 to 10, which is essentially unheard of these days. Nowadays they would both cut deliverables in half, while quadrupling costs, for an 8x increase in "cost per x" for the project overall.
I recently read How Big Things Get Done by Bent Flyvbjerg, and he has actual data by project type on things like the Boston Big Dig or California high speed rail, and it was pretty eye opening how so many of our large projects are impossible boondoggles now.
Related, Tesla is backing off continuing to push the gigapress forward, at least temporarily. The rationale is largely that the target vehicle platform is no longer a priority, but also there is a significant upfront capital investment matched with a long timeline to perfect it. I wonder if, given the push to industrialize but with modern manufacturing tech, there is a place for the national labs to get involved? https://www.reuters.com/business/autos-transportation/tesla-retreats-next-generation-gigacasting-manufacturing-process-2024-05-01/
One rumor I read about the biggest difficulty using giga-presses is creating the dies, which often require many tweaks to get working right. The key enabling technology here was 3d printing large sand molds quickly to iterate through many designs. There's so many challenges that go into making these things work. Investing in infrastructure like giga-presses again would probably be a better way to support American industry long term than throwing more money at unprofitable businesses. Just saying!
In your "Heavy press program" table, the the location of the two Wyman Gordon forges is misspelled as "Worchester." The name of the Central Massachusetts city is "Worcester."
- BYD is still welding stamped steel sheets together, and BYD prices are actually lower than Tesla's
- Tesla is no longer planning single unitary castings for future vehicles
- unitary cast aluminum bodies could have a cost advantage for certain production numbers, like 200k cars, but dies for casting wear out sooner than dies for stamping steel, and as soon as you need to replace them the cost advantage is gone.
- robotic arms are flexible and stamped panels can be used for multiple car models, and if you already have robots and panels you can use from discontinued car models, the cost advantage is gon
- bigger castings have more problems with warping and voids. Also, a bigger casting with the same curvature from warping has bigger position changes
Great article as usual. I have a question. I was reading that once jet powered engines were introduced not only did it make planes fly faster but also made them safer. How is that possible?
I don’t have any figures to hand, but one factor is that jet turbines are much more reliable than piston engines. Turbines have 2ish orders of magnitude fewer moving parts than piston aero engines, more straightforward cooling and lubricant arrangements, and being constant rotation vs reciprocating have longer fatigue life for components. Kerosene turbine fuel is also less flammable than avgas, which probably contributes to safe operations too.
Great article. Thanks.
How about a follow-up about the methods used in the actual Pressing or Extruding process?
The Press needs something to give the part a desired shape. I'm assuming some kind of strong/hard pattern/plate above and below the material to be pressed. What are they made of? How long do they last? How are they made? Is the forming done as a single slow procedure or multiple strikes? I've seen videos of steel railway wheels being forged from rectangular blocks... How is this different?
In Extrusions, how do they get the hollow centers from solid material? How is butted tubing made?
This is a wonderful article. I had no idea that these very heavy presses needed such government support, and it's rather depressing that further progress in this important field of manufacturing will be, in practice, made by the Chinese.
I think most of us think of manufacturing as a boring, low-tech field with doesn't require much R&D, and certainly doesn't require programs to be sponsored by the government. How wrong we are.
Fantastic Article. That was fascinating.
When I was pre-teen living near Cleveland, Ohio in the 1960s, Alcoa had a Bring-Your-Family-to-Work day. I got to see the 50,000 Mesta forge press in action. It stood several stories high and also extended several more floors into the ground. When it moved, it appeared like slow motion. Not what I expected. The top die would descend for several seconds. When it impacted the work piece, there would be a shutter...like an earthquake and a low frequency rumble. You could feel this in every building in the area.
Dad graduated from the University of Illinois in Champaign in what was a new field at the time: Metallurgical Engineering. The school also had a large press on campus.
Thanks for the article.
Another thing worth noting about these projects was the tremendous speed at which the presses were designed, fabricated, and put into production. Seems like 3 to 5 years, which included deciding on and obtaining the site, building the ancillary structures, and building and assembling the presses themselves. I don't know if this was the norm back then, but now it seems to take 3 to 5 years just to get the contracts signed for any significant effort.
That was exactly my thought. "Ah, this was back when we actually had the minimum competence and state capability to actually accomplish moderate stretch goals. Those were the days, weren't they?"
Not only that, but they kept the "cost per press" about the same even after dialing it back from 17 to 10, which is essentially unheard of these days. Nowadays they would both cut deliverables in half, while quadrupling costs, for an 8x increase in "cost per x" for the project overall.
I recently read How Big Things Get Done by Bent Flyvbjerg, and he has actual data by project type on things like the Boston Big Dig or California high speed rail, and it was pretty eye opening how so many of our large projects are impossible boondoggles now.
Related, Tesla is backing off continuing to push the gigapress forward, at least temporarily. The rationale is largely that the target vehicle platform is no longer a priority, but also there is a significant upfront capital investment matched with a long timeline to perfect it. I wonder if, given the push to industrialize but with modern manufacturing tech, there is a place for the national labs to get involved? https://www.reuters.com/business/autos-transportation/tesla-retreats-next-generation-gigacasting-manufacturing-process-2024-05-01/
One rumor I read about the biggest difficulty using giga-presses is creating the dies, which often require many tweaks to get working right. The key enabling technology here was 3d printing large sand molds quickly to iterate through many designs. There's so many challenges that go into making these things work. Investing in infrastructure like giga-presses again would probably be a better way to support American industry long term than throwing more money at unprofitable businesses. Just saying!
Small correction, Worcester, MA doesn't have an h in it.
Yes, just like Worcestershire! No ‘h’ in either spelling or pronunciation.
In your "Heavy press program" table, the the location of the two Wyman Gordon forges is misspelled as "Worchester." The name of the Central Massachusetts city is "Worcester."
Very interesting article. These heavy presses are very impressive pieces of machinery.
Very informational and educational. Thanks for sharing!
Hey, id be curious to hear your opinion on the following article, arguing that the giga press was actually not the right method for Tesla.
https://www.lesswrong.com/posts/nZcb9TTyneFbxiDav/the-giga-press-was-a-mistake
boost and I would want to know also.
to sumarise the post argument:
- BYD is still welding stamped steel sheets together, and BYD prices are actually lower than Tesla's
- Tesla is no longer planning single unitary castings for future vehicles
- unitary cast aluminum bodies could have a cost advantage for certain production numbers, like 200k cars, but dies for casting wear out sooner than dies for stamping steel, and as soon as you need to replace them the cost advantage is gone.
- robotic arms are flexible and stamped panels can be used for multiple car models, and if you already have robots and panels you can use from discontinued car models, the cost advantage is gon
- bigger castings have more problems with warping and voids. Also, a bigger casting with the same curvature from warping has bigger position changes
Great article as usual. I have a question. I was reading that once jet powered engines were introduced not only did it make planes fly faster but also made them safer. How is that possible?
I don’t have any figures to hand, but one factor is that jet turbines are much more reliable than piston engines. Turbines have 2ish orders of magnitude fewer moving parts than piston aero engines, more straightforward cooling and lubricant arrangements, and being constant rotation vs reciprocating have longer fatigue life for components. Kerosene turbine fuel is also less flammable than avgas, which probably contributes to safe operations too.
so good