Don't give up on the SPIF... the way the pros do it is they alter the shape of the tool path to account for the built up tension. So the raw formed sheet (before trimming) doesn't look like the end product, but after cutting it settles into place. Most of them use in-house developed simulations to do this automatically, but you might be able to get by with guesstimate work. I also suspect annealing might be helpful (although I don't know enough about Ti to know if that's realistic).
Did you see smarter every day’s video about robot sheet forming? The material gets pinched between two robot fingers that move together in three dimensions. Fascinating.
Yes. Anneal the metal before cutting the part from the sheet. You have to relieve the residual stress. Heat treat after cutting the part out to get strength properties back.
You would need an algorithm that changes the rpm or feedrate according to the point of touch of the ball to achieve perfect rolling (when possible). Very interesting remark.
Im fairly certain that I saw a video here on the tube a few years ago of someone 3D-printing all kinds of dies and shapes to then stamp/cut iron sheet with. They were far more durable than I would have ever imagined. 3D-printing is fairly low-cost compared to having to mill a new shape every time =) Cool video and interesting idea you have =)
Got to wonder if after you finished the forming op, what would happen if you ran it through a heat cycle to normalize it and release some of that tension. Because you're kind of work-hardening as while you form it, right? But stamping doesn't really have that problem, and that's kind of what you're doing is stamping but really slowly, so idk. SmarterEveryDay had a video on this I think last summer or fall. Did a walk-through on a place trying to make this viable for mass production. Lots of good details there.
@@RussellMakes yeah just leave it in the fixture before the cut-out cycle, and hit it with a torch, you can tell it's annealed once it changes color. Also use an endmill with a less aggressive spiral angle, and do a spiral-in to cut through the aluminum first before you hog out around the perimeter.
That's a cool idea. It was fun watching those pieces form. I know the material was fairly thin, but any concerns about the loads being put on the CNC spindle bearings?
Welcome to the issues with forming Ti. It work hardens while cutting and requires huge forces for bending. I think the SR71 guys wanted a 24000ton press to form their Ti sheets
Maybe look at the tool-paths used in metal-spinning? Because metal-spinning also does a lot of mechanical deformation with stretching and compression of the sheet.
There is a special grease for high pressure aplication. As i know there are particles in the grease to prevent the grease from being pushed away due to the pressure.
Someone already mentioned it, but starting from the inside might be better. When you start on the outside you are making a work hardened ring that does not want to yield so you are trying to stretch a progressively smaller section of soft metal thus increasing the stress and distortion. Maybe you could start in the middle and do about half then move to the outside and finish it or vice versa, or start in the mid point. Maybe doing something like this you could induce counteracting forces thus limiting distortion.
i'm pretty sure that it is the way, the problem is - regular toolpath isn't going to help, you basically could not push lower point down to intended depth from the beginning, not without catastrophic failure at least, so you have to slice toolpath into multiple layers from lowest to highest points, each of which starts from same Z0, but at the same time pushing previous toolpath result deeper
You beat me to it. I was watching one of the dedicatted industrial machine videos doing this and thought it would an interesting experiment on my CNC router. I planned to use a hard plastic tip and some spray lube.
I've seen the more advanced version of this where they have two end effect robots using styluses from each side of the sheet. Not sure if I recall them doing titanium but they can do some pretty deep draw pulls in the sheet with that technique. But then again not everyone has millions of dollars for that equipment and software (monitors pressures, paths optimization and where on the stylus they are pressing to prevent galling and tearing etc). Pretty good results here though.
I wonder if titanium is work hardening. You may want to hit the titanium with a torch to see if that relaxes the internal stress before parting the formed part from the surrounding material.
Maybe you could add a post forming heat treat step prior to cutting the part out with endmill. Some stress relief cycles and annealing and possibly tempering prior to machining if free could resolve the intense internal stresses causing the movement while machining. Obviously depending on the particular metal the process would be different and the results will vary. Also given your setup is capable, I might consider milling a form type die or (soft jaws) to support the part from the bottom of the fixture. Bassically the reverse profile of the part being formed. It would likely improve surface finish as well as create the support needed to go thin without blowing through the material.
if u could have the tool touching from both side like the big machines ive seen do it, but most of those are using big robotic arms to do the work but still how hard would it be to make a machine that would have two spindles one facing up and one down an wouldnt the bottom just be a mirror of the top. i love this kinda stuff but dont know how to do it at all. i used 3d printers an manual mill an lathe but ive always wanted to get where i could do cnc mill an this is just something else that makes me want to learn it so i can play with this kinda stuff but i dont know if i'll ever get there or not
What if you indexed one corner to stop the sheet from spinning and then use the clamps to only prevent vertical movement? This would allow the sheet to be pulled in. Also, set the tool path to build up the veritcal shaping in relative layers where the "deepest" part is displaced first.
WOuld it be possible to do this in multiple passes? they would have to be multiple passes on each z layer I assume. You might get less springy if you make the deformation in more steps? What would happen if you heated the TI up after the process do you think you could reduce the tension?
Very interesting idea. I would have thought about hydroforming into a a 3D printed die. Still have to deal with spring back but you might get steeper walls with that technique because there should be less work hardening.
tbh, if you 3d print the molds you might be as fast as with this SPIF setup. There are some great papers out there doing just that, even having one sided dies that press into rubber. And dies that are made from a bunch of individual shapes so that you don't have to make the whole die again and again. Write me a PN if you are interested 😉I think i have the pdfs somewhere in my dropbox.
When cutting out the part you should try leaving a 1.5 diameter wide onion skin, then come back in and cut out the part with tabs. When the sheet springs, it wont matter because you are only cutting through foil thick material. Probably easer said than done with sheet metal.
I think if you annealed the titanium by hitting it with a torch, and then setting a cover over it for a few hours before cutting it out, it would hold its shape better.
Instead of using the end mill, just run the tool harder and tear the material. It would be interesting to see if the tension in the material would be different using that method.
I've been wanting to try something similar, but I want to do it with an air hammer, mounted in a C style frame. then the sheet metal would be pinched with each hammer blow. I think its worth giving a shot, though I have quite a few other projects I am working on at the moment, so it will be a while...
Could you introduce a fixture that allows you to flip and hit it from the other side without releasing it? Might significantly stiffen up the surface to put a feature in a feature, help keep things planar/within shape. Incredibly cool, looking forward to see where you're headed!
This is an awesome process! Now I want to try too! Can you talk about the tool paths and how you made them? I. Guessing it's just 3d paths from fusion with specific step downs, but I'd love to know more!
You're going to want to try to match the rotation speed of the ball with how fast it's traveling across the materials. Also, if you're trying to cold form without any heat, grease would be your best bet.
1.) A friend of "this old tony showed" that he annealed aluminium. paint the alu with a sharpy and heat it up, until the sharpy is gone. 2.) maybe the metal is shifted to one side only, because you do not change the direction of the tool path. Maybe it is better to intermediate change from clockwise to anit-clockwise toolpath.
The other thing you might play with is running the spindle at high speed; specifically because it will heat up the material kind of like a flow-drill. 🤔
Work hardening and material flow, titanium is very difficult to work with this process. I would try sheets of copper or brass to practice the process, easy to anneal.
No idea. I would guess pretty low considering the sheet I’m working with is pretty thin. The only one that heated up at all was the stainless piece which was about 0.038” thick.
A $400 shoe horn? Also, I kinda fast forwarded through this but you could use a “pitch” as a backing material that would help vibration and spring back.
Wouldn’t you want to start forming from the deepest point? You wouldn’t hammer form a part from the outside in as you need to draw the material from as far away as possible. That’s why you’re getting those ridges and cracks.
So, I don't agree... You need to include that deformation as a part of your process. You should be able to precompute the expected bends in it and reverse it out of the generated gcode... that way, when you do release it, it will bend to your desired shape. Pre and post heat treating it will help relieve those internal stresses before you try and cut it free as well.
@@LocnavLivoc I think so. I ended up calculating the diameter of the forming tool at the point of contact with the work piece and determined what rpm to spin based on my feed speed so the tool would theoretically be closer rolling along the surface instead of slipping and relying completely on the grease or oil. Seemed to help the surface finish.
It looks like you are using the wrong type of toolpath. If you have it, try the scalloping equal stepover toolpath such as in Mastercam. the x,y and z step overs are kept equal.
Don't give up on the SPIF... the way the pros do it is they alter the shape of the tool path to account for the built up tension. So the raw formed sheet (before trimming) doesn't look like the end product, but after cutting it settles into place. Most of them use in-house developed simulations to do this automatically, but you might be able to get by with guesstimate work. I also suspect annealing might be helpful (although I don't know enough about Ti to know if that's realistic).
Did you see smarter every day’s video about robot sheet forming? The material gets pinched between two robot fingers that move together in three dimensions. Fascinating.
Came here to say this.
Titanium can be annealed to relieve stress. Also, some sheet-metal forming processes (deep drawing) work best with lanolin as the lubricant.
Yes. Anneal the metal before cutting the part from the sheet. You have to relieve the residual stress. Heat treat after cutting the part out to get strength properties back.
The stress comes from the work hardening that this process causes.
As others mentioned, annealing the metal would help a lot, before and after forming. Might have to re-temper depending on your application too.
You would need an algorithm that changes the rpm or feedrate according to the point of touch of the ball to achieve perfect rolling (when possible). Very interesting remark.
Im fairly certain that I saw a video here on the tube a few years ago of someone 3D-printing all kinds of dies and shapes to then stamp/cut iron sheet with. They were far more durable than I would have ever imagined. 3D-printing is fairly low-cost compared to having to mill a new shape every time =)
Cool video and interesting idea you have =)
Not on titanium
you're probably thinking of one of "stuffmadehere" 's videos
Got to wonder if after you finished the forming op, what would happen if you ran it through a heat cycle to normalize it and release some of that tension. Because you're kind of work-hardening as while you form it, right? But stamping doesn't really have that problem, and that's kind of what you're doing is stamping but really slowly, so idk.
SmarterEveryDay had a video on this I think last summer or fall. Did a walk-through on a place trying to make this viable for mass production. Lots of good details there.
Hmm, might have to try that. Thanks for the idea.
@@RussellMakes yeah just leave it in the fixture before the cut-out cycle, and hit it with a torch, you can tell it's annealed once it changes color. Also use an endmill with a less aggressive spiral angle, and do a spiral-in to cut through the aluminum first before you hog out around the perimeter.
Go Green!
That's a cool idea. It was fun watching those pieces form. I know the material was fairly thin, but any concerns about the loads being put on the CNC spindle bearings?
Welcome to the issues with forming Ti. It work hardens while cutting and requires huge forces for bending. I think the SR71 guys wanted a 24000ton press to form their Ti sheets
You could try to start from the inside out.
I think you need some backup from behind to take the tool forces, like a medium density open cell foam.
Maybe look at the tool-paths used in metal-spinning? Because metal-spinning also does a lot of mechanical deformation with stretching and compression of the sheet.
There is a special grease for high pressure aplication. As i know there are particles in the grease to prevent the grease from being pushed away due to the pressure.
Someone already mentioned it, but starting from the inside might be better. When you start on the outside you are making a work hardened ring that does not want to yield so you are trying to stretch a progressively smaller section of soft metal thus increasing the stress and distortion.
Maybe you could start in the middle and do about half then move to the outside and finish it or vice versa, or start in the mid point. Maybe doing something like this you could induce counteracting forces thus limiting distortion.
i'm pretty sure that it is the way, the problem is - regular toolpath isn't going to help, you basically could not push lower point down to intended depth from the beginning, not without catastrophic failure at least, so you have to slice toolpath into multiple layers from lowest to highest points, each of which starts from same Z0, but at the same time pushing previous toolpath result deeper
is there a reason to go from the outside in, instead of inside out?
Would a form on the backside help at all? Might support the edges better and keep the pieces from warping as much.
You beat me to it. I was watching one of the dedicatted industrial machine videos doing this and thought it would an interesting experiment on my CNC router. I planned to use a hard plastic tip and some spray lube.
Very cool to see you try this at home. I might have to have a go.
I've seen the more advanced version of this where they have two end effect robots using styluses from each side of the sheet. Not sure if I recall them doing titanium but they can do some pretty deep draw pulls in the sheet with that technique. But then again not everyone has millions of dollars for that equipment and software (monitors pressures, paths optimization and where on the stylus they are pressing to prevent galling and tearing etc). Pretty good results here though.
I wonder if titanium is work hardening. You may want to hit the titanium with a torch to see if that relaxes the internal stress before parting the formed part from the surrounding material.
Would it be better and more consistent to just oil or wax the whole sheet? Uniform, thin film, for a more uniform result?
That's amazing. I had no idea you could do that on a regular old mill.
Maybe you could add a post forming heat treat step prior to cutting the part out with endmill. Some stress relief cycles and annealing and possibly tempering prior to machining if free could resolve the intense internal stresses causing the movement while machining. Obviously depending on the particular metal the process would be different and the results will vary. Also given your setup is capable, I might consider milling a form type die or (soft jaws) to support the part from the bottom of the fixture. Bassically the reverse profile of the part being formed. It would likely improve surface finish as well as create the support needed to go thin without blowing through the material.
if u could have the tool touching from both side like the big machines ive seen do it, but most of those are using big robotic arms to do the work but still how hard would it be to make a machine that would have two spindles one facing up and one down an wouldnt the bottom just be a mirror of the top. i love this kinda stuff but dont know how to do it at all. i used 3d printers an manual mill an lathe but ive always wanted to get where i could do cnc mill an this is just something else that makes me want to learn it so i can play with this kinda stuff but i dont know if i'll ever get there or not
What if you indexed one corner to stop the sheet from spinning and then use the clamps to only prevent vertical movement? This would allow the sheet to be pulled in. Also, set the tool path to build up the veritcal shaping in relative layers where the "deepest" part is displaced first.
WOuld it be possible to do this in multiple passes? they would have to be multiple passes on each z layer I assume. You might get less springy if you make the deformation in more steps?
What would happen if you heated the TI up after the process do you think you could reduce the tension?
Very interesting idea. I would have thought about hydroforming into a a 3D printed die. Still have to deal with spring back but you might get steeper walls with that technique because there should be less work hardening.
tbh, if you 3d print the molds you might be as fast as with this SPIF setup.
There are some great papers out there doing just that, even having one sided dies that press into rubber. And dies that are made from a bunch of individual shapes so that you don't have to make the whole die again and again.
Write me a PN if you are interested 😉I think i have the pdfs somewhere in my dropbox.
When cutting out the part you should try leaving a 1.5 diameter wide onion skin, then come back in and cut out the part with tabs. When the sheet springs, it wont matter because you are only cutting through foil thick material. Probably easer said than done with sheet metal.
I think if you annealed the titanium by hitting it with a torch, and then setting a cover over it for a few hours before cutting it out, it would hold its shape better.
Instead of using the end mill, just run the tool harder and tear the material. It would be interesting to see if the tension in the material would be different using that method.
The manufacturing direction of the sheet metal (grain?) might be influencing the outcome, it does when bending at least.
I've been wanting to try something similar, but I want to do it with an air hammer, mounted in a C style frame. then the sheet metal would be pinched with each hammer blow. I think its worth giving a shot, though I have quite a few other projects I am working on at the moment, so it will be a while...
Good idea but I feel this may kill your CNC!
Could you introduce a fixture that allows you to flip and hit it from the other side without releasing it? Might significantly stiffen up the surface to put a feature in a feature, help keep things planar/within shape.
Incredibly cool, looking forward to see where you're headed!
Very cool. Thanks for sharing!
Could reversing the tool path for a light preforming run help?
This is an awesome process! Now I want to try too! Can you talk about the tool paths and how you made them? I. Guessing it's just 3d paths from fusion with specific step downs, but I'd love to know more!
Yes exactly, I ended up using 3d adaptive on 0.030” slices of the shape I wanted. Also did spiral which worked great for the cone and sphere.
I wonder if switching between center out and outside in when changing z
You're going to want to try to match the rotation speed of the ball with how fast it's traveling across the materials.
Also, if you're trying to cold form without any heat, grease would be your best bet.
Maybe somehow anneal the sheets before removing them from the workholder to ease out forming stresses? Is that even possible?
1.) A friend of "this old tony showed" that he annealed aluminium. paint the alu with a sharpy and heat it up, until the sharpy is gone.
2.) maybe the metal is shifted to one side only, because you do not change the direction of the tool path. Maybe it is better to intermediate change from clockwise to anit-clockwise toolpath.
I've never done it but it looks to me like you might need to roughly form the center of the sheet first instead of trying for 1 finish pass.
maybe graphite grease is the way to go
and annealing it before hand
The other thing you might play with is running the spindle at high speed; specifically because it will heat up the material kind of like a flow-drill.
🤔
Work hardening and material flow, titanium is very difficult to work with this process. I would try sheets of copper or brass to practice the process, easy to anneal.
Any idea what the axial forces are compared to normal milling tool pressures?
No idea. I would guess pretty low considering the sheet I’m working with is pretty thin. The only one that heated up at all was the stainless piece which was about 0.038” thick.
Which diference with press mold??
Cool project
Have you tried annealing after you finish?
Could you make the ball offset a little so that it wobbles and turn up the RPM to “hammer” the metal a little?
A $400 shoe horn? Also, I kinda fast forwarded through this but you could use a “pitch” as a backing material that would help vibration and spring back.
Very cool dude!
Maybe try running a current through the metal to heat it when forming?
No clue about spif, but why would you do this instead of forming with a press and mold?
Molds take production time, can cost a lot, and take up shop space. If you're doing a one off part or proof of concept, spif may be better.
Need to use grease, not oil and anneal super often. Titanium work hardens super fast.
Wouldn’t you want to start forming from the deepest point? You wouldn’t hammer form a part from the outside in as you need to draw the material from as far away as possible. That’s why you’re getting those ridges and cracks.
What did you used ti geberate rhe tookpath?
Have you tried it while flooding it?
Are you looking at any way of supporting the underside? I believe the originator of this technique has...
So, I don't agree... You need to include that deformation as a part of your process. You should be able to precompute the expected bends in it and reverse it out of the generated gcode... that way, when you do release it, it will bend to your desired shape. Pre and post heat treating it will help relieve those internal stresses before you try and cut it free as well.
try annealing the sheets after to normalize it and get rid of the internal stress
My boss says you could try molybdenum disulfide for lube, and about 50 rpm.
No idea if that’s good advice 🤷♂️
3d printing for stamped parts, essentially
question: does the rotation direction matter?
@@LocnavLivoc I think so. I ended up calculating the diameter of the forming tool at the point of contact with the work piece and determined what rpm to spin based on my feed speed so the tool would theoretically be closer rolling along the surface instead of slipping and relying completely on the grease or oil. Seemed to help the surface finish.
@@RussellMakes thanks
Use Militec ( or another metal conditioner ) for lub
Using your current toolpath looks the same as 3D printing layers
hydroforming?
Back with rubber?
Try Lessing the pressure as it’s going down
Now you need a really huge mill and you can make a car 🙂
Can’t pinch the material around the edge like that. Need a bigger ball tool and tighter passes. Need thicker material.
It looks like you are using the wrong type of toolpath. If you have it, try the scalloping equal stepover toolpath such as in Mastercam. the x,y and z step overs are kept equal.
3D print dies. It works
Did you happen to watch the smarter everyday video about pinch sheet metal forming?
Temper first?
make some subwoofer cones
I’d be afraid to do that with m spindle
@@shaunybonny688 why? The force is pretty low. It’s just sheet metal
pretty SPIFfy
Seems an expensive way to make spoons.
Interesting 🤔🤔😊
maybe start in center and work out - backwards of what doing now
relax it with a heating cycle before you cut it out
👍🏻👏🏻👏🏻👏🏻👏🏻👏🏻👏🏻👏🏻👏🏻
Also.... Please say you're designing custom cod pieces.
Titanium, huh? So... gonna build a SR-71?
😜
You should wear safety glasses. Carbide can shatter.
Just change the amount of machining you're doing on the Titanium! Machine more shallow than the aluminum and it should spring to where you want it.
Very hard on the machine, tools male and female and a PRESS.
Have you thought about investing in a CNC water jet? It would cut those parts out with ease and you wouldn't break end mills.
Your clamp angle sucks.
very brave and very foolish of you to apply the oil while the machine is moving
Oh Shut up. not a single person wants to hear you crying about what another man should or should not do.