Body buck making techniqiues, preferences?
Discussion
Steve_D said:
Has anyone costed the CNC route?
For most the possibility of modelling the car in CAD is a non starter but a body buck can be measured by laser or probe and converted into the cad model albeit just as an outline.
The advantage is that only one side of the buck needs to be sculpted as the resultant digital image just needs to be mirrored.
Only having to produce half a buck may make the required savings.
Yes, I'm doing that with Aeon Sportscars with a remake of the Aeon GT3.For most the possibility of modelling the car in CAD is a non starter but a body buck can be measured by laser or probe and converted into the cad model albeit just as an outline.
The advantage is that only one side of the buck needs to be sculpted as the resultant digital image just needs to be mirrored.
Only having to produce half a buck may make the required savings.
Before I built my own CNC machine, we were quoted something like £6,000 for a rough cut (this is ~25mm smaller than the body, and you place clay over it to finish it off before taking the mould), and £25,000 for a fine cut (where you would be able to more or less take a mould off of the machined work).
We're not exactly sure how much it'll cost with my own machine, but at a guess, it's about £3,000 for a fine cut using foam designed for the job. We could probably do a fine cut using something that may interact with the mould (we'd need some form of isolation layer) for about £1,000. The accuracy I want is to the nearest 1mm (the machine can do ~0.06mm in X and Y axes, and 0.01mm in Z), but when you're cutting at that accuracy, the time to cut goes out the window (my aim is for 7 day cutting time, but that is at a 1mm accuracy).
The machine has cost me around £3,500, but it can be used for other things as well.
tribbles said:
Yes, I'm doing that with Aeon Sportscars with a remake of the Aeon GT3.
Before I built my own CNC machine, we were quoted something like £6,000 for a rough cut (this is ~25mm smaller than the body, and you place clay over it to finish it off before taking the mould), and £25,000 for a fine cut (where you would be able to more or less take a mould off of the machined work).
We're not exactly sure how much it'll cost with my own machine, but at a guess, it's about £3,000 for a fine cut using foam designed for the job. We could probably do a fine cut using something that may interact with the mould (we'd need some form of isolation layer) for about £1,000. The accuracy I want is to the nearest 1mm (the machine can do ~0.06mm in X and Y axes, and 0.01mm in Z), but when you're cutting at that accuracy, the time to cut goes out the window (my aim is for 7 day cutting time, but that is at a 1mm accuracy).
The machine has cost me around £3,500, but it can be used for other things as well.
I presume you would do the body in sections, and that the issue is how far the machine moves along the main axis between each cut? Is the idea to do the final cut on a hard surface than can then be sanded to remove the ridges left by the tool?Before I built my own CNC machine, we were quoted something like £6,000 for a rough cut (this is ~25mm smaller than the body, and you place clay over it to finish it off before taking the mould), and £25,000 for a fine cut (where you would be able to more or less take a mould off of the machined work).
We're not exactly sure how much it'll cost with my own machine, but at a guess, it's about £3,000 for a fine cut using foam designed for the job. We could probably do a fine cut using something that may interact with the mould (we'd need some form of isolation layer) for about £1,000. The accuracy I want is to the nearest 1mm (the machine can do ~0.06mm in X and Y axes, and 0.01mm in Z), but when you're cutting at that accuracy, the time to cut goes out the window (my aim is for 7 day cutting time, but that is at a 1mm accuracy).
The machine has cost me around £3,500, but it can be used for other things as well.
singlecoil said:
I presume you would do the body in sections, and that the issue is how far the machine moves along the main axis between each cut? Is the idea to do the final cut on a hard surface than can then be sanded to remove the ridges left by the tool?
Yes.It would be possible to do a really fine cut, but I suspect that the time would be astronomically long.
The machine itself is actually quite quick (I've had it cutting at 18m/min - professional machines are between 20 and 30m/min, and the fastest I've seen is 90m/min). Unfortunately that's in some cheap extruded polystyrene (wall insulation panels), and I'm guessing I'll only get around 5m/min in the material we're aiming at cutting (it's a lot harder).
We'll be using block sizes of around 600x1200x100; a very rough estimate for my design is that it'll be around 60 blocks - although not all of them are full blocks. Some of them can be really thin sections, so you'll find that only a small handful require a cut over the entire surface; most would be partial cuts, and that reduces time.
A frame would be made to hold the blocks, since there's no point in cutting blocks that have no cuts
tribbles said:
Yes, I'm doing that with Aeon Sportscars with a remake of the Aeon GT3.
Before I built my own CNC machine, we were quoted something like £6,000 for a rough cut (this is ~25mm smaller than the body, and you place clay over it to finish it off before taking the mould), and £25,000 for a fine cut (where you would be able to more or less take a mould off of the machined work).
We're not exactly sure how much it'll cost with my own machine, but at a guess, it's about £3,000 for a fine cut using foam designed for the job. We could probably do a fine cut using something that may interact with the mould (we'd need some form of isolation layer) for about £1,000. The accuracy I want is to the nearest 1mm (the machine can do ~0.06mm in X and Y axes, and 0.01mm in Z), but when you're cutting at that accuracy, the time to cut goes out the window (my aim is for 7 day cutting time, but that is at a 1mm accuracy).
The machine has cost me around £3,500, but it can be used for other things as well.
25k sounds a lot to me for a body buck. I would hope for something closer to 15-18k for a buck similar to the ginetta one on another thread, including finishing. How would the buck be constructed for that kind of money, and what would constitute a fine cut?Before I built my own CNC machine, we were quoted something like £6,000 for a rough cut (this is ~25mm smaller than the body, and you place clay over it to finish it off before taking the mould), and £25,000 for a fine cut (where you would be able to more or less take a mould off of the machined work).
We're not exactly sure how much it'll cost with my own machine, but at a guess, it's about £3,000 for a fine cut using foam designed for the job. We could probably do a fine cut using something that may interact with the mould (we'd need some form of isolation layer) for about £1,000. The accuracy I want is to the nearest 1mm (the machine can do ~0.06mm in X and Y axes, and 0.01mm in Z), but when you're cutting at that accuracy, the time to cut goes out the window (my aim is for 7 day cutting time, but that is at a 1mm accuracy).
The machine has cost me around £3,500, but it can be used for other things as well.
egomeister said:
25k sounds a lot to me for a body buck. I would hope for something closer to 15-18k for a buck similar to the ginetta one on another thread, including finishing. How would the buck be constructed for that kind of money, and what would constitute a fine cut?
I wasn't involved in the "negotiations" for it (which I think were just "I'll phone them up and get a rough idea"), so I don't know what the quality of finish was. Perhaps "quoted" may have been too strong a word...There may be differences if there's more sections to the car, or if there's complicated fiddly bits.
Seeing as we are noe discussion options of 25K what would a one off aluminium body cost?
The Whittney-Paine special posted somewhere on PH looked lovely as do the the Kirkham cobra bodies but I've no idea how much they cost.
For a one off you can drive the result and you can take molds of it. These two aspects combined might make it worth considering.
Does anyone have a clue what a one off would cost?
The Whittney-Paine special posted somewhere on PH looked lovely as do the the Kirkham cobra bodies but I've no idea how much they cost.
For a one off you can drive the result and you can take molds of it. These two aspects combined might make it worth considering.
Does anyone have a clue what a one off would cost?
tribbles said:
Steve_D said:
Has anyone costed the CNC route?
We're not exactly sure how much it'll cost with my own machine, but at a guess, it's about £3,000 for a fine cut using foam designed for the job. We could probably do a fine cut using something that may interact with the mould (we'd need some form of isolation layer) for about £1,000. The accuracy I want is to the nearest 1mm (the machine can do ~0.06mm in X and Y axes, and 0.01mm in Z), but when you're cutting at that accuracy, the time to cut goes out the window (my aim is for 7 day cutting time, but that is at a 1mm accuracy).The machine has cost me around £3,500, but it can be used for other things as well.
If you made the buck in smaller pieces could you use a much smaller, and therefore lower hourly rated, machine?
What is it that makes it that expensive?
cymtriks said:
tribbles said:
Steve_D said:
Has anyone costed the CNC route?
We're not exactly sure how much it'll cost with my own machine, but at a guess, it's about £3,000 for a fine cut using foam designed for the job. We could probably do a fine cut using something that may interact with the mould (we'd need some form of isolation layer) for about £1,000. The accuracy I want is to the nearest 1mm (the machine can do ~0.06mm in X and Y axes, and 0.01mm in Z), but when you're cutting at that accuracy, the time to cut goes out the window (my aim is for 7 day cutting time, but that is at a 1mm accuracy).The machine has cost me around £3,500, but it can be used for other things as well.
If you made the buck in smaller pieces could you use a much smaller, and therefore lower hourly rated, machine?
What is it that makes it that expensive?
The initial foam blocks we used were £25 each (1200x600x100), and somewhere between 30 and 60 of them are needed - which gives a cost of £750 to £1500. However, in bulk the prices would be lower.
The blocks we are looking at now (which we can take the mould off) is around 3 times the price, but is much firmer. This means the machine would need to go slower. The blocks are actually modelling foam, so are fine-grained, and you can even tap them (which is what we'll probably be using to hold them down).
The time cost isn't something we've calculated, but it shouldn't actually be that much because it doesn't require someone constantly watching over it. Someone would need to pop their head around the corner from time to time to make sure the machine hasn't gone wrong, or check if it needs the block replacing.
This is all experimentation for us at the moment - the only figures I have is how much it's cost us so far in development time and parts (which is about 10 weeks and £5000 [which includes an early prototype that didn't really work]).
The early prototype was originally designed to do a whole car at once - however, there are a number of problems with doing that:
1) You need to have a large space to put the machine in
2) You need to have huge blocks of foam
3) There's more wastage (unless you can make the foam into a rough shape)
4) It'll take longer (cutting tools have a finite depth, so you'd need to work your way down to it)
5) You really need a 5-axis machine so you can do sideways cutting
6) If it goes wrong, you may be looking at replacing the whole block (eek!)
7) The machine needs to be VERY sturdy
The prototype failed on points 5 (it was only a 3 axis machine) and 7 (it was very wobbly).
In using blocks in a 3D jigsaw, it's a lot easier to do, and we can partially simulate having a 5-axis machine (despite it being a 3-axis machine).
The software I've written takes the 3D object and works out the block arrangement. Blocks at the front and back are milled front and back; the sides are done side to side, and the rest are milled downwards. There are occasions where it can't do a particular type of cut - that would need to be done by hand.
andygtt said:
I would love to see that working, opens up a world of creating other components such as the dash etc.
It has been working: http://www.youtube.com/watch?v=Y9FvOffAIvc(Main site is http://www.tribbeck.com/steel/stlcutter2/ )
However, I partially dismantled it a month ago to add an additional ballscrew for the x-axis, and it's still in the process of being re-assembled (it's about 70 miles from home, which makes it a little harder for me to do it).
ETA: The video says it's running at 12m/min, but it's actually 15 in the Y axis, and 18 in the X (I'd forgotten that I had tweaked the speeds after the initial tests).
Edited by tribbles on Friday 20th November 13:15
Paul Drawmer said:
That's impressive.But also close for what we had for the Mk3 version
I like the fact that they're doing really rapid cuts in very cheap foam, coating it with some finer material, and then doing a fine cut on that.
I'd imagine their machine costs a little more than mine (I'm guessing between £500K - £2M).
andygtt said:
Body filler on wood is a bad idea, its what I have done and basically the filler distorts as it dries and the shape is constanty changing.
try to use something that doesnt shrink as it cures.
We used to spray them with epoxy resin to get them to settle and give some strength before putting filler over the top.try to use something that doesnt shrink as it cures.
PhillipM said:
andygtt said:
Body filler on wood is a bad idea, its what I have done and basically the filler distorts as it dries and the shape is constanty changing.
try to use something that doesnt shrink as it cures.
We used to spray them with epoxy resin to get them to settle and give some strength before putting filler over the top.try to use something that doesnt shrink as it cures.
Would the cheapest CNC approach be:
Make a car sized former and fill with cheap foam
Cut back (by CNC or by hand using station templates) to about 1-2 inches below the final surface
Cover with fibreglass to give a firm base for the finishing layer
Apply the finishing layer (car body resin probably)
Cut back by CNC
However...
All of this assumes that the CNC route is the best. I'm not convinced.
First you need to make a CNC machine. That's a big project in itself.
Then you need to learn how to use a CAD package to draw your car. A big learning curve.
Then you need to learn how to get the shape into the CNC machine (tool path software etc). Another learning curve.
Then you need to construct your buck out of lots of foam and filler.
The result will need hand finishing (The DP1 needed circa 200 hours of rubbing down to get a perfect finish)
Then you need to take the molds off this. This means another learning curve or bought in skill.
Now you make the panels. This means another learning curve or bought in skill.
That lot has to be almost as big a project as making the car!!!
I seriously wonder if a aluminium body shell might not work out cheaper for a prototype. You can drive it straight away and can take moulds off it if required.
Then of course there is always doing it the time honoured kit car way using lots of time and effort to do it by hand.
Make a car sized former and fill with cheap foam
Cut back (by CNC or by hand using station templates) to about 1-2 inches below the final surface
Cover with fibreglass to give a firm base for the finishing layer
Apply the finishing layer (car body resin probably)
Cut back by CNC
However...
All of this assumes that the CNC route is the best. I'm not convinced.
First you need to make a CNC machine. That's a big project in itself.
Then you need to learn how to use a CAD package to draw your car. A big learning curve.
Then you need to learn how to get the shape into the CNC machine (tool path software etc). Another learning curve.
Then you need to construct your buck out of lots of foam and filler.
The result will need hand finishing (The DP1 needed circa 200 hours of rubbing down to get a perfect finish)
Then you need to take the molds off this. This means another learning curve or bought in skill.
Now you make the panels. This means another learning curve or bought in skill.
That lot has to be almost as big a project as making the car!!!
I seriously wonder if a aluminium body shell might not work out cheaper for a prototype. You can drive it straight away and can take moulds off it if required.
Then of course there is always doing it the time honoured kit car way using lots of time and effort to do it by hand.
singlecoil said:
PhillipM said:
andygtt said:
Body filler on wood is a bad idea, its what I have done and basically the filler distorts as it dries and the shape is constanty changing.
try to use something that doesnt shrink as it cures.
We used to spray them with epoxy resin to get them to settle and give some strength before putting filler over the top.try to use something that doesnt shrink as it cures.
cymtriks said:
That lot has to be almost as big a project as making the car ... an aluminium body shell ...You can drive it straight away
Although I agree with you about the CAD/CNC route I think you're downplaying or underestimating the effort involved in crafting an ally body. Define 'straight away' - I'm not busy next weekend... 
You mean straight away after you've learnt how to be metalwork craftsman or after that, putting in the hundreds of hours actually making the body in question? Unless we're sticking to some seriously limited forms sculpturally. The Whitney-Paine you previously cited is quite complex and IIRC they did use some sort of 3d software in the process.
Whichever method is chosen this is not a small undertaking and one's choice may depend on other factors, such as whether this is this a standalone project or can investing extra time in learning new skills be offset against subsequent projects?
seansverige said:
cymtriks said:
That lot has to be almost as big a project as making the car ... an aluminium body shell ...You can drive it straight away
Although I agree with you about the CAD/CNC route I think you're downplaying or underestimating the effort involved in crafting an ally body. Define 'straight away' - I'm not busy next weekend... You mean straight away after you've learnt how to be metalwork craftsman or after that, putting in the hundreds of hours actually making the body in question? Unless we're sticking to some seriously limited forms sculpturally. The Whitney-Paine you previously cited is quite complex and IIRC they did use some sort of 3d software in the process.
Whichever method is chosen this is not a small undertaking and one's choice may depend on other factors, such as whether this is this a standalone project or can investing extra time in learning new skills be offset against subsequent projects?
I am suggesting that having a body made in aluminium might be easier/more cost effective than going down the self build and self taught CNC route.
Caterham did it this way with the Caterham21. An alloy body original was used as the test car and as the styling buck.
By "drive straight away" this is what I mean. Once the body is made you can drive off. Going the self taught CNC route you have to jump through a whole series of big expensive projects and climb several learning curves just to get to a buck which you still can't drive until you have jumped through another set.
Look at what has already been done on the project in question. Thousands of pounds spent, two machines built, a further machine required, models produced (which may well need revising), CNC tool paths created (which will need revising for the next machine) and still a whole lot more expense away from a driveable car.
It would be interesting to compare this with the cost of something like the Whitney-Paine car, done as a subcontract job. A whole Kirkham Cobra costs circa 35000 pounds. This includes the chassis, suspension and lots of other bits besides the bodywork but also reflects using the tooling many times over.
You could buy an awful lot of filler an wood even with some of the smaller figures quoted on here an if you want to keep things cheap an easy you wont go far wrong with birch faced ply,jelaton an filler.Several ways to do it but a firm wooden base just below the basic profile with the surface in a mixture of the items is cheap an easy an easily achieved at home in your garage with a few basic tools
The more work you put in the better it will be,dont cut corners an you will have an exellent pattern which is the start for a good mould which leads to a good part
The more work you put in the better it will be,dont cut corners an you will have an exellent pattern which is the start for a good mould which leads to a good part
Gassing Station | Kit Cars | Top of Page | What's New | My Stuff