CAD and CAM surfboards

General Discussion
21
Quote:

Those machines are decent for the price, though accuracy is not good .015". and speed is slow.

If we were to build a machine we could make an asome one, super fast and accurate, it would be more expensive but not by a whole lot, especially if we got input from people here.

While 0.015" definitely isn’t the rather fanatic accuracy one gets from the high end CAM machinery ( which is typically accurate to a few thousandths of an inch, not 15/1000 of an inch) , still, boards are finish sanded by hand, glassed by hand, hotcoated by hand, sanded by hand, glossed by hand…and , just checking, 10 sheets of 24 lb inkjet paper measure around 1/8" - 0.125", so the 0.015" cited is about the thickness of a piece of paper.

Anything more than this is, frankly, overkill. I’d think that getting much more accurate without One Helluva Heavy Machine would be difficult…and unnecessary. Who is gonna glass that close, hotcoat that close, sand that close and gloss that close?

Let alone expensive - R&D costs would be right up there, tooling up to make the things, etc.

Speed is slow- well, yes and no. Bear in mind that each pass it makes is gonna be full-depth, cutting to the finish profile and not several passes with a planer. More on this in a bit…

What else? Well, how many of these are ya gonna sell, so how many do you have to amortise your R&D costs on. 200? 500? 23? 2? How big is the surfboard industry? How many production shops can it support? With ( realisticly ) $20,000 and up machines? Versus machines that are more general and work in a bunch of different industries, thus having lower costs and prices? ?

Now, considering cutting speeds and producing things… there’s something they use in servers, those gizmos that Swaylock’s and most web sites are stored on, called RAIDs - Redundant Arrays of Inexpensive Devices.

Huh?

Yeah, what they are is storage. Instead of making a super big hard drive that Will Never Break Down, Ever, Really, We Mean That!!..which is probably impossible and horribly expensive if it was possible… they use a bunch of relatively cheap hard drives, operating in parallel, so that if one breaks down your data is still there on another one, so you swap out the busted one, or fix it, and you’re still in business. In fact, you never went out of business, though you may have slowed down a little.

Okay, so if production speed is important, X number of shaped boards per day beyond what one cheap shopbot can whittle…get two, or three, as that’s still cheaper than the hypothetical ‘surf special’. What’s more, if one breaks down, you’re not dead in the water, you still have one or more others to keep on plugging with. Your night guy, Jeff Spicoli, will have several machines to feed, not just one, so he’s not out getting stoned with his buddies between boards. Your investment is lower, your production is higher and your situation vis a vis production is a lot better.

And Shape3d not only does the shape but generates the tool paths and directly supports the Shopbot format? Well, for 500 euros, that’s a win.

Leastwise, that’s my take on it. I’d have to say it’s pretty much a solved problem.

doc…

22

The paper I measure measures, .0025, but I guess there are lots of different kinds of paper.

Any way a tolerance of .015 means one pass may be low by .015" and the next high by .015" so really it ends up throwing you off by .030" which isn’t a lot when it comes to surfoboard shaping I know. But put a little wear on that machine and that tolerance is going to open up rather quickly, you just can’t avoid it with a rack and pinion drive system like that.

When I machine a board I like to machine it so it is almost ready to glass, I don’t mind letting the machine run all day. With a tolerance like that it wouldn’t work for my needs. Generally I do step overs of .050", which is quite a bit where I come from. But with a quick rub with a screen it is done and I havn’t altered the geometry too much.

Some good points though doc.

I think the shop bot would be a great buy for someone just starting out in the CAM stuff. But I have found in my business that buying quality machines and tools pays off in the long run.

23

So what’s the philosophical difference between a mass produced board entirely made on a machine (including rails and concaves) and a pop-out??

I can understand the rough shaping that’s then finished by hand, or the commercial reality of needing a machine to make the standard models in volume, but what’s the difference between that and a pop-out??

Regards,

Matt.

24

You know, you got me thinking a bit…considering a vaccuum clamping system, deflection and all while machining… without getting into something like a full length vaccum support, contoured and all to support and hold everything rigidly, just how close can you machine something out of irregular foam in a machine tool like one of these gantry mills?

I mean, we’re not talking about carving something out of a billet of steel on a Bridgeport here ; with the support always right under the tool, a flat table, a flat surface resting on the table and several rigid hold-downs like those use, we’ve got a chunk of foam that can only be held down with a few vaccum setups to support it - isn’t the foam gonna deflect a little, given a fairly quick feed and a fair sized tool?

Yeah, you probably could program the software to account for that, but it’d be a beeyotch to do, and it’d have to be different for every blank and every stringer configuration and every foam weight…gah. I’d think there’s a certain point beyond which you just have to figure you’re not gonna get any closer…

Dunno, you’re a working machinist, what’s your thoughts…

doc…

25

kenz, did the 6" Z travel work for you or did you go with the custom Z mod?

I’m still going through all the required pieces with my uncle, but this shopbot machine is looking pretty good. It is about the accuracy I wanted to go for originally. It cuts a little bit slower than I had hoped for but I’m not sure how much faster you could even go before tearing the foam. If building your own machine turns out to be a huge saver I’ll make sure to post the specs on it.

26
Quote:

So what’s the philosophical difference between a mass produced board entirely made on a machine (including rails and concaves) and a pop-out??

I can understand the rough shaping that’s then finished by hand, or the commercial reality of needing a machine to make the standard models in volume, but what’s the difference between that and a pop-out??

Regards,

Matt.

The difference is a pop-out you are stuck with one design, because it is made from a mold that costs quite a bit of money to make. Plus they are not made from the same materials as standard boards, generally people consider the materials better, though there may be problems with flex.

A CNC milled board doesn’t have to be the same every time and it can be made from the same materials we have all grown to know and love. The great thing about CAD designed and CNC milled boards is you can take an existing board and make it 1/8 of an inch thinner or change the rail minutely, then take it out and see how it is different, then you learn what you like better, every board you can continue to fine tune your board.

Pop outs are great if it just so happens that you love one of the boards they make, then your stoked you know you can order 20 of that perfect board and they should all ride almost exactly the same.

27
Quote:

kenz, did the 6" Z travel work for you or did you go with the custom Z mod?

I’m still going through all the required pieces with my uncle, but this shopbot machine is looking pretty good. It is about the accuracy I wanted to go for originally. It cuts a little bit slower than I had hoped for but I’m not sure how much faster you could even go before tearing the foam. If building your own machine turns out to be a huge saver I’ll make sure to post the specs on it.

you can cut the foam as fast as the machine will go, so long as the cutter is spinning somewhat fast. except the stringer you have to be carefull. Please let me know what you come up with.

28

I agree with droptrow. Machined boards are where custom boards truly lie. With software like shape3d you can digitally recreate any board by hand in less than 20 minutes and then fine tune it to perfection.

29

Mine came with 7" Z travel, which allows me to machine boards with a 5" nose rocker no problem. I think 6" Z would be too little for this much nose rocker, because you need to allow at least an inch for the router bit. I’ve only done a nose rocker more than 5" once. I just had t0 it machine up to 5" and then hand shaped the last few inches. If I found myself doing bigger rockers I wouldn’t hesitate to get a new z-axis.

As far as speed, you can run it at 4" per second. That’s pretty fast. It goes through foam like butter. There is no tearing. The finish is perfect after a light sanding. On the ShopBot forum some have mentioned running it at up to 6" per second through foam. I haven’t been in that much of a hurry to try it.

30

If going too fast isn’t an issue in the foam, then the feedrates can definitely be stepped up. Shape3D is setup to cut the stringer first and at a slower feedrate.

droptrow, you’re the machinist, how fast is too fast in foam? some of the linear slides I’ve been looking at are capable of 12Gs acceleration and feedrates of a couple meters per second. This is obviously overkill for surfboard shaping. What would you say is the ideal feedrate, acceleration, and spindle speed combo? How awesome would that be to pop out one side of a blank in less than a minute?!?!

31

The sepcs for the PRT say that the maximum xy feedrate is 4" / sec. Did the ones running it at 6" / sec have to modify the drivers?

32

Does anyone know how much the CNC version of Shape3D runs for without the machine?

33

wow I read acouple words in this thread out of a few random entries and kept scrolling wow holy robotix batman… Issac Asimov is peering in from behind the veil…will Fred Saberhagen sign in at the front desk? I for one can provide only 1/2 models as my accuracy is asymetric,as well as far inferior to the Inky poo…me and John Henry will be waitin down by the railroad spur…he’s the one with the 12# hammer in each hand and I’m the one with the crookit foil form and the smile wearing trunks and a smile covered with dust… ambrose…stop by for a chat

34
Quote:

I mean, we’re not talking about carving something out of a billet of steel on a Bridgeport here ; with the support always right under the tool, a flat table, a flat surface resting on the table and several rigid hold-downs like those use, we’ve got a chunk of foam that can only be held down with a few vaccum setups to support it - isn’t the foam gonna deflect a little, given a fairly quick feed and a fair sized tool?

First off I don’t use as big a tool as the production guys, usually I use a 3/8 or 1/2" mill, so I can get down in channels and cut wings and things better. But no I don’t notice any deflection when cutting through foam, its like cutting air. I use 6 vacuum cups stagered down the board and 3 sets of leveling posts to keep it level and still. also I clamp the sringer at the nose and tail. that system holds it pritty rock solid. The problem I do have though is once the stringer is cut off the bottom the whole thing wants to bow a little, but I have solved this by first cutting the top but leaving a little material, then cut the stringer first on the bottom, then the foam, then flip it back and take the top to size. Way over kill probably, but its just the way I work and I don’t mind because these boards are for me and I enjoy doing it. Overall, how I make my boards is probably over kill I admit, but at the same time you’d also be suprised at how little it takes to alter a boards ride. I am just experimenting, I want to take precision to the extreme and see what it does for me. Is it really helping or should I just go back to hand shaping because Im not getting results. If I hadn’t started doing this I probably never would have known that by taking material off the stringer it releases stress and causes the rocker to change. I think that is a good thing to know.

35
Quote:

If going too fast isn’t an issue in the foam, then the feedrates can definitely be stepped up. Shape3D is setup to cut the stringer first and at a slower feedrate.

droptrow, you’re the machinist, how fast is too fast in foam? some of the linear slides I’ve been looking at are capable of 12Gs acceleration and feedrates of a couple meters per second. This is obviously overkill for surfboard shaping. What would you say is the ideal feedrate, acceleration, and spindle speed combo? How awesome would that be to pop out one side of a blank in less than a minute?!?!

One thing you might want to think about for Z travel.

Ideally in your machine you would be able to flip the board by having your nose and tail stringer clamps spin off a fixed, known axis. This will help by being able to know where the board is. then in your machining program you just flip the board on that same axis and you are set to machine the bottom. but to do this you will need like 12" of z travel at least, and you would need your fixture to be able to retract pritty far, but if I made a machine that is how I’d do it I think, unless someone else had a better way to flip it.

I don’t know of a speed limit for such light foam, the main limiting factor for cutting stuff is heat, does the friction cause heat to build up and destroy the cutter or the material. and from what I have seen I don’t think heat is a problem. So I think go for the 12G’s if you can afford it. But make sure you have a router that spins at at least 32,000 rpm just to make sure the cutting edge isn’t having to cut through 1/16" or more of foam, that possibly could cause tearing. Of course I can’t guarantee anything because I have never seen it done.

I imagine with 12G’s your going to have to have a seriously heavy machine. I have worked on a large CNC router before and we had it cutting 400 inches per minute through plexiglas with a 36K RPM spindle, it cut fine, but when the thing went to a sharp corner the whole machine jumped, and this was a heavy machine. We eventually started slightly rounding all corners, and that took care of it.

Losos, where are you looking at the linear slides? I’d like to check them out.

36
Quote:
Any way a tolerance of .015 means one pass may be low by .015" and the next high by .015" so really it ends up throwing you off by .030" which isn't a lot when it comes to surfoboard shaping I know. But put a little wear on that machine and that tolerance is going to open up rather quickly, you just can't avoid it with a rack and pinion drive system like that.

When I machine a board I like to machine it so it is almost ready to glass, I don’t mind letting the machine run all day. With a tolerance like that it wouldn’t work for my needs. Generally I do step overs of .050", which is quite a bit where I come from. But with a quick rub with a screen it is done and I havn’t altered the geometry too much.

Sounds like you want to shape boards to aerospace tolerances.

For everyone else, let me clarify a few things about my own experience. I’m not a machinist, but to me tolerance means that when you ask the machine to cut to a certain depth and then measure it afterwards, the tolerance is the difference of what you wanted and what you got. There are a lot of variables involved in tolerance such as how well you assembled your machine to fluctuations in the temperature. I took the time to assemble my machine as precise as possible, and I consistently get between .005" and .008" tolerances.

And it’s inaccurate to suggest that the tolerance changes so dramatically from one pass to the next. The temperature would have to drop or climb dramatically with each pass. If I were to machine a flat section, it doesn’t fluctuate in depth with each pass, it’s completely flat. There is no sanding required.

With the ShopBot you could run it with an 1/8" bit and have it step over .050", and you would get a board that would require almost no sanding. But it would take all day. It 's just not worth it. Most machine shaped boards are machined with a 1" rounded end bit and a 1/2" to 3/4" step over. Then they are lightly sanded with a disk sander using a super soft pad.

37
Quote:
The sepcs for the PRT say that the maximum xy feedrate is 4" / sec. Did the ones running it at 6" / sec have to modify the drivers?

As I recall, they just did it with the standard step motors. They were just testing the limitations of the machine.

38
Quote:

Sounds like you want to shape boards to aerospace tolerances.

For everyone else, let me clarify a few things about my own experience. I’m not a machinist, but to me tolerance means that when you ask the machine to cut to a certain depth and then measure it afterwards, the tolerance is the difference of what you wanted and what you got. There are a lot of variables involved in tolerance such as how well you assembled your machine to fluctuations in the temperature. I took the time to assemble my machine as precise as possible, and I consistently get between .005" and .008" tolerances.

And it’s inaccurate to suggest that the tolerance changes so dramatically from one pass to the next. The temperature would have to drop or climb dramatically with each pass. If I were to machine a flat section, it doesn’t fluctuate in depth with each pass, it’s completely flat. There is no sanding required.

With the ShopBot you could run it with an 1/8" bit and have it step over .050", and you would get a board that would require almost no sanding. But it would take all day. It 's just not worth it. Most machine shaped boards are machined with a 1" rounded end bit and a 1/2" to 3/4" step over. Then they are lightly sanded with a disk sander using a super soft pad.

I think we are talking about this too much, but whatever, its all a fun healthy debate.

Yes, why not shape boards to aero space tolerances? Why does aerospace shape to aerospace tolerances? Is it because they work better that way?

Well, generally tolerance means it is the maximum fluctuation from the intended measurement, not just one measurement. So take and cut a 5" circle at maximum speed and take some calipers or a micrometer set to exactly 5.008" and see if it will go all the way around. Then take and set them at to 4.992" and see that it doesn’t fit on any where. That should tell you if your tolerance is within .008", except for the Z axis of course. Please let me know what you find, I am skiptical but I would love nothing more than to be proven wrong because it would be nice if that machine did fit my needs, for that price you can’t beet it.

Be carefull when you say “completely flat”.

For me a machine taking all day to cut a board is fine because I will just set the machine in the garage, turn it on in the morning and come back after work and flip the board and cut the rest while I sleep, I would rather the machine take the time than me spend my precious time sanding something that could have been done more accurate by the machine. I relize some people may have more than 1 board to do a day and that option wouldn’t work for them. But for me that is how I would do it.

39

Im no expert but i used to work in a C&C machine shop. I was told that one of the main reasons the aerospace industry machines parts to very high tolerences is that if you have a part spinning at extremly high RPMs that tiny or what seems like a tiny flaw is the difference between the part surviving and the part blowing the whole thing apart.

40

Yes like I said, it works better.

Who is to say that a small bump on the bottom of a surfboard doesn’t blow the whole ride of the board apart? or exactly how much effect it does have on the ride? Plus tolerances play a hug factor in a lot of things. if parts don’t fit together properly they can wear, or if tolerances are off they just won’t fit together at all. It is not only at extreems that tolerances come into play. But generally things work best when they are made to the highest tolerance possible, its just the cost has to factored in.