If you got free expert help for your shop & shopbot, what would it be?

[henrik_o]

Ok, first of all, I want to say that this is all very preliminary. Nothing might come of it, but since the ball looks like it's moving I figured what the hey, why not post about it.

I'm about to become a 'member' of the Robot Valley Project (RVProject) here in Sweden. The aim of this project is to aid productivity and cnc/robotics adoption in member companies and to do original public research and development in these fields. My part is still very preliminary, but I've had a general consultant over, for no less than three hours, and he was very excited about our Shopbot, it was overall an extremely pleasant meeting.

As I understand it, a projects manager will sit down with us to discuss what we have, where we want to go, what is limiting productivity, and so on. From that, one or several company-specific projects might emerge. These projects will be a collaboration between the member company and a small staff (which really might mean just one or two) students plus a supervisor at our local industrial college. If expert attention is needed, this can be provided at a significant discount to us (up to 50% paid by the RVProject). If it is detemined that a project involves technology not specific to the company, I think they can throw more resources at it, though I'm not sure I understood that part fully.

An example of a project that we discussed was for a local industrial painting shop. The company bought a 'vanilla' industrial robot, and the RVProject fitted it with the necessary hardware to make it an autonomous 'painting cell', not just the hardware on the robot and the systems feeding to it, but the whole arrangement in the shop, how to load and evacuate, and so on.

So, with the caveats as per above, what might be interesting projects to propose?

If you look at what you have now, and you were given free time by a gang of students plus some less expensive time by experts to work on something, what would it be?

I have a couple of ideas, but to start the thread on a completely open note, what would be your pet project?

[gerryv]

Tailor-made luges, skeletons, snow-boards, etc.; complete or components or molds for these?

[jhicks]

As I look at our process, clearly the cutting time is no longer the focus of our attention. Rather the 10 minutes it takes to remove the parts, identify/label the pieces that were cut for later use/assembly, clear the dust, reload and start again.
With an average cut time ranging from 15 to 45 minutes, this extra 10 minute runload/reload cycle burden eats up over an hour a day of otherwise productive machine capacity.
It would be very nice to have some simple and cost effective ideas to reduce this by 50% or more.
Sounds like a great opportunity to enhance your process and work with some talented students. Cant wait to hear what comes of it.

[henrik_o]

Gerald,

That's an excellent example of what they could help with, but it's not something we do and wouldn't be very interesting for the general shopbot user group, I imagine.

[henrik_o]

Jerry,
quote:As I look at our process, clearly the cutting time is no longer the focus of our attention.
That's my situation as well. Good to know I'm not the only one. I know I could optimize certain files better, and so on, and hey there's always that toolchanger you secretly pine for, but realistically speaking, what is hampering productivity is the load-evacuate-sort-clean-load cycle.

quote:It would be very nice to have some simple and cost effective ideas to reduce this by 50% or more.
Bingo. This is exactly what I've been having in mind.

Me and the general consultant went over our application and a few ideas emerged at this early stadium;

1. A vacuum hoist system married to pneumatically actuated stops on the table; the hoist would need to be solid and stable, and all controls (for the hoist function as well as the stops) collected on one remote. This should be connected to a good logistics system for handling sheets up to the point where they are to be loaded.

2. An automatic evacuation system. Basically, one could use actuators to lower a bar on the back of a PRS gantry, and then drive/push the machined sheet off the table using the x axis motion of the shopbot. Lots of issues involved, but I think it could be done. Possibly integrate some cleaning function.

3. An automated cleaning system. Basically lowering a bar on the back of a PRS gantry, using x axis motion to sweep/clean the table. Debris would first be caught by gradually finer brushes (the push motion -- it could be collected by a hood at the end of the table), and behind the brushes a 'sealed' zone where compressed air is used to agitate finer dust into a hood served by the DC system.

4. A more advanced dustfoot/dusthood. Ideally, one that has better general dust evacuation properties, is more ergonomic and capable of cleaning all toolpaths even when using negative (downcut) cutters. A combination of compressed air and vacuum is probably needed.

5. ... (there was some more odds and ends) ...

We went over how we do it now, talked and talked about it, a few good and easy suggestions for things we can do right away came up, and we've exchanged some emails since. A very pleasant experience.

I don't know if this will actually recieve a green light, but then again I don't see why not. Most of the above is pretty generic stuff that would suit their 'public' agenda quite well, it can be used on 3-axis cnc's regardless of brand. We will see, in any case.

Comments and suggestions very much welcomed.

[bcammack]

Would it be impractical to create a narrow, slotted plenum at one end of the Shopbot table, tied to the dust collection system? You could grab a push broom and just push the remaining debris to it from the other end of the table.

That might expedite post-job cleanup of the table.

[jeffreymcgrew]

Brett: that's what we did sort of. We put a dust collector run to one end of the table, where we've also enclosed around the spoilboard to the frame (boxed in the whole of the table if you will). It's just a 4" round hole, but you pull the plug out, and then just sweep towards it, and it all goes up the pipe and into the collector...

[Gary Campbell]

Henrik..
RE: #4 I have been using that combination.... good luck! It seems hard to get anything out of the cut groove more than 1/2" below the surface... up or down spiral.
Gary

[bstern]

Henrik and Jerry - do you have some cameras set up in my shop? This is exactly what my biggest frustrations are. I currently am producing 4 sheets per hour and know how I can get to 5 with some basic changes to machine and code. At 5 sheets an hour a large percentage of time is now taken up by labeling removing parts and cleaning and reloading.

With melamine all these issues are big and add parts cleaning.

During the sheet unloading I thought I could use the output of the vacuum as a blower to power an 4' wide air knife. This could be done in a box attached to the DC. The blower is quite powerful but only about 200 cfm. The DC port would only need about 400cfm to ensure no dust escapes at the input output slots.

Another approach would be to clean up better during the cuts. So this makes a better dust foot with some sort of compressed air agitator a big thing as well. With my PRS I think there would be enough power to load and unload at the same time. You could use compressed air inputted into the vac system to reduce friction. This would require much better chip removal than I am achieving now.

[jseiler]

(While a little off topic, this may be of some interest.)

If you've not read anything about production engineering or the Toyota Production Method, it may be a good time to take a good look at other bottlenecks in your production process (since you're likely to resolve some big ones). One particularly accessible book, written in novel format, is "The Goal."

http://www.amazon.com/Goal-Process-Ongoing-Improvement/dp/0884270610

It goes through the theory of constraints pretty well, showing that once you clear one production time constraint, another will appear. The book may be common sense, but I found it to be very enlightening, myself not having much experience in production/industrial engineering.

John

[bstern]

Thanks John. I have heard of the book but, never got around to it. I think I will order a copy.

Had a Production and Ops professor at Wake Forest named Herb Schutte.
He gave me his self proclaimed - Herb's curse.
I will never look at a product and not think about how it is made. He also made me look at bottle necks and do an analysis of how much it's going to cost to solve it. You need to look at the gain between the first bottle neck and, as you said the proverbially next one, to see the value of solving the one at hand. Some times you need to look 2 or 3 ahead and average them out.

Right now as I just started my business my bottle neck is sales!!!!

As soon as I get that moving forward I will put resources on production again.

[henrik_o]

Brett,

quote:Would it be impractical to create a narrow, slotted plenum at one end of the Shopbot table, tied to the dust collection system? You could grab a push broom and just push the remaining debris to it from the other end of the table.
I think this is a very easy modification/add-on that would be quite helpful. The problem as I see it is that using a manual broom (at least on our 3m+ table) will clean 90-95% of the debris, but it won't actually clean the table. You'd still need to blow it off with compressed air to really get it clean, and that's a big no-no for health reasons.

The system I dream of would have an automatic broom, ejecting debris into a vertically oriented DC hood at the end of the table, but after the broom part moving along the X would be a sealed area where compressed air is used to agitate debris/dust into a hood connected to the DC system or a Fein/Festool vacuum.

I'll whip up some pics of what I have in mind since it's kind of hard to convey in words.

[henrik_o]

Gary,

quote:RE: #4 I have been using that combination.... good luck! It seems hard to get anything out of the cut groove more than 1/2" below the surface... up or down spiral.
How did you set it up? I agree that directing compressed air into packed paths doesn't really work, you'll get some out but it tends to just pack the rest even tighter.

To my thinking, the air vector must agitate as the cut happens; if packing occurs, blowing at it won't agitate more than the surface, agitation must stop packing from occuring. I don't know if it is possible to do this, but it would be interesting to hear if anyone's been experimenting.

If the dust foot is at a constant height over the material, rather than traveling along with the Z, it should be possible to have four nozzles in the dust foot, x+; x-; y+; y-; all aimed at the actual center of cut.

[henrik_o]

Bob,

quote:Henrik and Jerry - do you have some cameras set up in my shop?

No, but it figures.

quote:During the sheet unloading I thought I could use the output of the vacuum as a blower to power an 4' wide air knife. This could be done in a box attached to the DC. The blower is quite powerful but only about 200 cfm. The DC port would only need about 400cfm to ensure no dust escapes at the input output slots.
Do you mean as in cleaning the sheet going in/out, or cleaning the table? Both?

quote:With my PRS I think there would be enough power to load and unload at the same time. You could use compressed air inputted into the vac system to reduce friction.
How do you mean, enough power in the PRS to do what? On the compressed air into the vacuum table, I think some industrial systems use something called inverse/reversed vacuum, 'floating' sheets off the table.

[henrik_o]

Anyway, here are some very rough outlines of how an automated cleaning system could look like. It would be in the form of a bar, or a "wing", that would normally be situated towards the back of the Y gantry, actuated by pneumatic cylinders attached to the inside of the sideplates on the gantry.

Here's a sketch of it in the "up" position:


7278

And in the "down" (active) position:

7279

A cross section of the functional elements;


7280

All in all, a pretty complicated system. Not something I would undertake myself, at this stage at least. If I'm given free time by a gang of students, who knows... ?

[Gary Campbell]

Henrik...
We found about cleaning the cut groove as a side benefit of using an air nozzle for bit cooling while cutting plastic for a local sign co. I think that getting higher flow through the dust foot helps eliminate table cleaning more though.
Here is a post on the air nozzle:
http://www.talkshopbot.com/forum/messages/15719/27082.html?1201653294
Here is the one for the "hacked" dust foot:
http://www.talkshopbot.com/forum/messages/28/27056.html?1201061315
Here is a pic of them working together:

7281
Gary

[chiloquinruss]

"large percentage of time is now taken up by labeling removing parts" What if there was an inkjet printer head on the spindle frame that would make a lettering pass prior to the cutout phase. That way all the parts would be labeled, just the 'pick up the pieces' time would remain. I have seen industrial inkjet print heads that could print legibly from 2 or 3 inches away and they printed lettering on brick, unfinished fence lumber, even the side of a semi trailer. Russ

[henrik_o]

Gary, that's a very snazzy dustfoot indeed, both functionally and aesthetically: it comes very close to what I've envisioned as an 'ideal' dustfoot. Great work, and very inspiring!

Where I differ is that I'd like to eject to the side of the hood rather than to the back. Something like the following sketch (again, in an ideal world, where there aren't Y motors and Dual Z's killing lateral space);


7282

Coming back to the compressed air issue, I experimented a little during work today. It seems that a ~3mm nozzle from a distance of ~75mm and operating at ~2.5 bar, aimed directly at the center of cut at approx. 45 degrees and aligned to the cut path, provides enough agitation to prevent packing from occurring when onepassing 19mm MDF at 100mm/s with an 8mm downcut bit.

This is a bit more pressure required than I would have liked. If one were to construct a quad-nozzle set (x+;x-;y+;y-) that's an awful lot of airflow, even if one allows for slightly lesser pressure in each nozzle due to them working in tandem x/y.

Anyway, a quad-nozzle on the 'ideal dustfoot' as I imagine it would inject air something like this;


7283

I really like the LOC-LINE setup you have; I've ordered some nozzles and manifolds from them and I'm looking at installing some blowers very much like you've done here on our line borer.

Now that I'm writing this I've already thought of some ways to fiddle around with my ideal dustfoot dream, so I'll get right back to that

[bcammack]

I'm still trying to find the time to fab up that annular ring that sits in the hole where the tool passes through and fires compressed air downward and inward at the bit at 45° increments.

The articulated conversion I did to the standard dust foot works a treat, too.

http://www.talkshopbot.com/forum/messages/17997/25658.html?1195138158

[handh]

Brett,

Really like your design, do you have anymore pictures of it in the mounted position on the Bot. Also when are you going to sell a kit, or would you be interested in offering a little more information and would you be willing to cut out some of the parts for others on your waterjet.

[bcammack]

Thanks, Jeff.

I really do need to bring in my digital camera and take a couple shots of it installed.

The waterjet is usually busy cutting countertops out of slabs of engineered quartz, so I have to sneak in little projects like this when I can. I'm just one of the minions here, so I don't have much latitude for taking us into new side lines.


I'm happy to share the DXF file that lays out the "sleeve" and the two little brackets for the dust foot. I can't help but think that it could easily be cut out on the Shopbot in 6061 1/8" aluminum like I did with the waterjet. The 1" strip brush and the 16" drawer slide came from Grainger.

Is there somewhere here that the file could be placed?

[henrik_o]

Brett,

That's a neat design. I second the call for pics!


Did some more experimenting today. The injection of compressed air really is a big deal, since not only does it prevent packing; it also "seals" the toolpath when doing a straight x/y run -- this is something brushes or flaps won't do well afaict, they'll seal the top of the material but not the actual toolpath where all the action is in terms of the debris vector.

[bcammack]

Oh, I suppose I could upload as an attachment, duh...


The dashed line is reference for the programmer to the dimension of the raw material.

I have the parts nested for the narrow kerf of the waterjet.



CAD of sheetmetal for the dust collector mods.

7284 (24.8 k)

[Gary Campbell]

Henrik..
Nice design..I forgot to note on my earlier post, we were able to get the chips to blow clean in a few cases with our single blower. We found that when we had enough air to clear the tool slot, we also had enough to lift the part from the spoilboard. Luckily for us, we seldom have to worry about table cleaning times as we dont process a large number of sheets in a day.

Another thought: When you introduce a certain volume of compressed air into the vacuum chamber created at the cutting area, will it decrease the vacuum differential to the outside? This may lower the efficiency of the duct collection.

For now we will stick with the new dust foot and just use the blower when needed for additional cooling.

Keep us informed, Gary

[handh]

Thanks Brett,

[3imoh]

I use air at times to help blow out packed kerfs. One thing I have thought about exploring was to set up some sort of adjustable interval timer with a solenond to "puff" the air on and off at a steady rate. Might be more efficient than just having a constant blast of air, though I am not sure.

Any thoughts?

-howie

[richards]

Howie,
I was thinking the same thing as I read through this thread this morning. A simple '555' timer chip could generate the timing intervals (astable mode). Just Google '555 timer' to find a lot of information on using that chip. A solid state relay could be driven from the 555 and the solid state relay could drive a 120VAC pneumatic valve.

[henrik_o]

Gary,

Thanks, it's an 'ideal state' dream but anyway, you bring up some very pertinent points;

quote:We found that when we had enough air to clear the tool slot, we also had enough to lift the part from the spoilboard.
Yes, this is a concern, though I feel that careful nozzle positioning should hinder it; in any case since you bring it up, one of the points of having several nozzles each producing medium airflow would be to act against that.

quote:Another thought: When you introduce a certain volume of compressed air into the vacuum chamber created at the cutting area, will it decrease the vacuum differential to the outside? This may lower the efficiency of the duct collection.
This is a very important point, and one of the things I had in mind when I wrote "This is a bit more pressure required than I would have liked [...] that's an awful lot of airflow".

I don't think it's possible to know without experimentation, but I will say that I believe it will mandate a very good seal on the material top: probably a very short and dense brush that's spring loaded.

Howie and Mike,

I'm not sure I follow. It sounds intriguing, but besides limiting airflow, what would be the function of it?

Edit: Yay, 300 posts!

[richards]

By shooting short, frequent bursts of air you would reduce the load on the compressor and still have excellent chip removal, assuming that the on/off times were set correctly.

With an ON time of about 0.75 seconds and an OFF time of about 0.75 seconds, the air blast should be about to keep the 'slot' chip free. Having burst times that long would not over-heat the pneumatic value, particularly if you use a Solid-State-Relay with zero-crossing detection (the most common type of SSRs).

Typically, I use the MAC brand of pneumatic air valves, which have withstood years of this type of service without failure.

Using a constant air-flow would be easier, but my little 5-hp air compressor would have a hard time keeping up with that kind of flow.

[dirk]

Hey guys I'm all for automation, but a real easy way to increase your cycle time is to put a 1/4 inch piece of mdf on top of the bleeder board. Once your parts are cut slide all parts and 1/4 mdf off table to a cart or work area, your machine will be clean and ready, just put another 1/4 inch sheet on machine,load material and start cutting. You can process your parts while the next sheet is cutting, and recycle the 1/4 inch spoilboards,

[henrik_o]

I thought I should update this thread to say that I'm just back from a conference where this project was discussed, and at the evening dinner I was seated along with on the one hand the project point man for technology plus a consultant from one of the world's largest makers of industrial robots.

So, things are moving along (albeit at a rather glacial pace) and I had the opportunity to discuss many of the issues raised above with those two gentlemen.

If anyone has more ideas and input, it's welcomed.