How accurate is your machine?

[knight_toolworks]

I ma trying to figure out how much the average shopbot is deflected by cutting force.
say you cut a 2" square 1/2" deep with a 1/2" bit climb and normal how accurate would the cut be?

[harryball]

I can't speak for everyone, but I normally measure my accuracy in 100ths of an inch. If I push very hard (cut fast) I can be off 1/10 of an inch. If I cut sanely I can see 5/100ths or so. I spent a good deal of time tweaking.

If I use a 3/8" cutter at 2 ips in 1/8" hardboard I don't have anything able to measure the error.

I've often wondered what "typical" is but I don't think there is a single answer. Too many variables are invovled. So... what is the absolute best accuracy that the shopbot is capable of with a 1/2" bit cutting 1/8" thick stock with no load?

/RB

[Gary Campbell]

Guys...
My comparison is slightly different, but nonetheless, real world. Cutting 3/4" plywood for cabinet parts, 3/8 compression @6ips, climb/conventional with onionskin (.72 deep in 1st pass), there is approximately .025 -.030 difference between the 2 passes. Since the second pass is under little or no load, I assume that the deflection, which is more than acceptable, is in the first (climb) pass. I have seen little or no increase in this "watermark" moving up to 8ips, but will report back, as next job will be cut at 10ips. Also note that when cutting at 4ips or less there seems to be an unmeasurable amount.
Gary

[tmerrill]

After replacing pinions and upgrading to the 4G board a few months ago, I went through a complete alignment check of my PRT 4896. Since I do mostly small, individual projects, I test cut small squares and circles in both hardwoods and corian. Measuring the results with a 6" digital caliper I was getting results within 0.002". This was with a 1/4" downspiral bit. This may have been due to bit deflection, but it was accurate enough that I choose not to spend any more time on it. If it was, I have to believe that a 1/2" bit wouldn't deflect under my test conditions.

Tim

[gene]

Whats the purpose of trying to cut that fast? I cut cab parts (not lately with this economy) at 6idp and the results are satisfactory for me . Sure i could cut faster but i would rather have better quality cut than finishing a minute sooner. If the time frame is that close then thats too much pressure on me and i think i'll pass on that job.
How many people drive their vehicle as fast as it will go every where they go? I guess i am more of let the tool work at a pace that i am happy with and doesn't tear up stuff . I think i'll go fishin tommorrow...

[Gary Campbell]

Gene....
Ahhhhhh, let me count the ways....
1) Because we can? or at least to find out IF we can.

2) Because time is money. Less time means less cost, which will convert to either more profit and/or ability to lower price. A lower price per unit may mean the difference between getting jobs or writing proposals that others are beating.

3) An impromptu survey shows that the average price that large production shops charge to cut eCabs nested sheets is around $25, depending on a few other factors. IF it takes an hour to cut these sheets, then there will be no sense to own a ShopBot for this purpose. If you can cut 8 an hour, then there is good reason to justify the use/purchase of the machine. The real numbers, of course, are somewhere in between, but are in fact directly related to what speed that you can move or cut.

4) To find the limits of our machine in an effort to quantify its maximum production value or ROI. To weigh cost versus production in an effort to see if adding a ShopBot can have a better ROI than an $80K single machine.

Be aware that cut quality is a given in this situation. No one will outsource to anyone that cannot get plywood parts to within .010. Why should they? The "big iron" shop that I visited recently cuts them at 12ips single pass, which is 4 times as fast as I can. (6ips X 2 passes)

Tim...
That would be a good test, although I would like to see it done with larger pieces to allow machine to "come up to speed". If, in fact, the deflection is in the bit, then the 1/2" bit should eliminate the problem, but if the deflection gets larger, then the problem is in the physical or mechanical parts of the machine.
Gary

[gene]

Gary,
My point exactly. If you have that much work lined up then get 2 bots for the money as a big iron machine. is the cut @ 12 ips better than the bot cuts? As for me now we do not have enough work to keep an old man with a pocket knife buisy much less try to keep up with the payments on an 80,000 machine. If you purchase a machine that is capable of cutting @ 12 ips and the cut quality is there then go for it. I knew that the bot wouldnt do that when i purchased it and i still think that its a good piece of shop equipment to have. I dont have that kind of demand for panel processing , i'm just a little tad pole in a great big frog pond and i aint working myself that hard just to croak.

[rb99]

I would look into AXYZ machines before deciding...

Maybe things have changed but the really solid heavy duty construction made them popular in sign shops years ago, and from their web site, they are still on top of their game.

I wonder how much they are?

RIB

[henrik_o]

Gene, getting a second bot is great and all but while that doubles machine capacity it also doubles cycle time & resources. You now have two machines to feed and evacuate material from, clean the table, load files, maintain, tool up, etc. There is a very good reason to try to max your capacity with the lowest amount of cycles spread over the lowest amount of machines.

Anyway, as Gary said, I think it basically comes down to the enthusiast and/or competitive angle. For those that bid jobs where big iron is in the competition, getting as much out of your existing investment as is possible can make/brake good paying and typically quite large jobs. For those that manufacture and sell volume products, reduced machine time coupled with reduced cycle time is also important since those are not fixed costs.

For me, so far it comes down to the enthusiast angle. There are some things on the horizon that might change that, but for now it’s about getting to know the limits. I haven’t looked a lot at it, but some light experimenting, yes.

Assuming that the machine is well in order and maintained my finding is that on our Embla (A PRS Alpha) it almost entirely comes down to rigidity in the Z. Between the thumb and index finger, I’d say about 80% of force-induced deflection on the mechanical machine parts comes down to this alone: it is the first areas to receive those forces and unlike the others there’s not that much rigid materials to absorb it. Since Embla is a tall lady (14” Z) we learned quite a bit about this during the first months.

So, yes, raising the table is afaict the most important thing one can do to reduce flex in high speed scenarios. There is of course a case of diminishing returns, and one needs a bit of leeway for thicker than usual materials anyway, but raising it up as much as is practically possible is what I’d do if I get a job where reduced machine times are essential.

[jerry_stanek]

Yes you have two machines to maintain but while one is cutting you can clean the table, load files and do all the other stuff. You still have to do that with one machine so if one unit runs twice as fast it would still take longer.

[Gary Campbell]

Gene...
As Henrik states, there is no easy answer without a lot of research. With the price of the new Bots pushing $35K, 2 of them are close to the same price as the "big iron". Cycle time, especially with alternating machines, is not the easiest thing for me to figure as cutting time, place,pick and clean vary a lot. Depending on the machine, maintenence may be higher on 2 machines, yet another factor.

Like you, I also didnt think I was getting a 12ips machine, but I am getting beat up at the time clock with the workarounds for off sized parts. Like Henrik... we have been diligent trying to find the exact source of what I like to call: "Lateral force, generated by the bit, not fully contained by the machine" Many write this off as bit deflection, but my testing shows that 1/2" bits have the greatest part size errors when cutting over 6ips.

Jerry...
Do the math... 12ips is 4 times as fast as a 6ips times 2 passes. And yes, on/off and clean could be happening on one machine during the cutting time of the other. Thats what started the mental conflict. Conflict gets even larger if machine doesnt run on SB3, which is near and dear to me.
Gary