PDA

View Full Version : How to compensate for cast acrylic varying thickness?



Riki Potter
07-05-2012, 9:47 PM
I've been working with 3mm acrylic alot lately and I've started to have problems, due to the fact that it's cast (extruded isn't an option) the sheets vary in thickness, often from one end to the other, sometimes up to 0.5mm either way.
When I'm making anything with slots this is messing with things, we had to throw out an entire bowl because none of the pieces fitted together!
I realise I could just make the slots 3.5mm but that would end up making every joint loose for any of the non-outlier parts. How do you guys get around this?

Also it's probably worth noting that I'm making decorative bowls that don't require any glue so tight slots are necessary in some places.

Dee Gallo
07-05-2012, 10:13 PM
Low tech suggestion for what it's worth: use a nail file or sandpaper to make them fit. Acrylic is pretty soft, I've used micropads to soften sharp edges or whatever.

Glen Monaghan
07-05-2012, 11:17 PM
A lot depends on how you are designing/assembling things, but in general you want to make joints dependent on the edges you cut rather than on the too-variant thickness of the material. For example, people commonly try to make the width (as opposed to length) of a cut slot friction mate to the uncut face of a tab, which is problematic with cast acrylic because you can't count on the thickness of the tab.

If you want to ensure a friction fit, you need to rely on the contact between cut edges (such as the ends of tabs and ends of slots) rather than between cut edges (the sides of a slot) and uncut faces (the face sides of a tab). At left in the picture is a typical rectangular tab meant to fit into a rectangular slot. As long as the slot's end E is at least as wide as the maximum possible tab thickness and you cut the slot sides S at least as long as the length of the tab's face F, you can insert the tab into the slot. The loser the fit, the easier the assembly. Too loose and the joint won't hold by itself. However, tighten the tolerances too much trying for a good friction fit and insertion of a rectangular tab into a rectangular slot gets nigh impossible.

Since you can't control tab thickness, you can't rely on slot width to ensure a good friction fit; instead, rely on slot length and tab length. That is, instead of relying on the friction fit between face F of a rectangular tab and side S of the slot, rely on the fit between tab ends C and slot ends E. To make insertion easier while better controlling friction fit, make the actual tab slightly shorter than the slot but then add small bumps to the cut ends C of the tab as shown to the right in the picture. The extra width added by the bumps should make overall tab length a "delta" more than the slot length. The appropriate delta depends on how tight of a fit is desired and what materials are used. Harder and more brittle materials will need smaller bumps and tighter tolerances, while softer and more flexible materials can tolerate larger bumps with looser tolerances. This method works reasonably well with acrylic.

Open-end slots (slots that are on the edge of the material) can't use this method because one of the ends E is missing. Consequently, they have to rely on friction between the sides of the open-ended slot and a face. Depending on the material characteristics and dimensions, you might be able to put bumps on both sides S of the open-ended slot to accommodate varying tab thickness. The width of the slot has to be great enough to accommodate the thickest tab, while the size of the bumps has to be big enough to grip the thinnest tab... AND the material has to have enough give or flex or plasticity to insert the thickest tab without splitting the slot apart when the tab is inserted through the open end of the slot (as opposed to straight in from the side, where you wouldn't get the ramp effect of the bumps). This approach is touchier with acrylic because acrylic slots can crack if the fit is too tight.

-Glen

Riki Potter
07-05-2012, 11:44 PM
Dee, I didn't realise acrylic was fairly easy to sand/file, although it's not the best method for a production line we're aiming for, I'll keep it in mind to one-offs.

Glen, what a fantastic amount of help! I understand what your saying and I've actually done a couple of slots like that before, just didn't realise it was a better all-round approach I guess. You've essentially answered the question perfectly.
Thankyou so much for taking the time to explain it, I love finding out helpful people like you exist :)

Rodne Gold
07-06-2012, 1:47 AM
Extruded is a lot better re thickness tolerances - in the order of 2% , cast can vary + and - 10-20% over a sheet. So if its possible to make the item in extruded , I would suggest you do - hovever it is way more prone to stress cracking than cast especially if you glue with solvent based adhesives.- bear that in mind
IRO cast, Some mnfgrs are better than others , try find one that does cast with lower tolerances. We do not accept cast materials if the thickness tolerance is above 10%. If you are doing production , compensating and messing to get a good fit will cost you a lot more than finding better materials , even if it is a little more pricey.

Greg Facer
07-06-2012, 10:49 AM
Another option could be to mill down the tabs to a consistent thickness with a router, a simple table mounted router with fence would work. Unless it's clear, you won't see the edges (shouldn't look too bad anyways). You'd have to even out to the thinnest part of the material or thinner.

An added step for sure, but still a production type solution. Glenn's suggestion is great, but would work better with items interlocking a few ways, otherwise the joint might still work loose with the thickness varation, as I assume any of those slot and tab type projects are liable to do.

I'm working on slot and tab type projects too (or will be once I have a working machine), but am leaning towards "hooks" for the tabs for more positive locking together. In those designs, you still need bumps like Glen illustrates for the variations and for some tension on the joint.

Good luck, and thanks Glen for the advice! I will try that method too, especially for iitems I can glue after assembly.

Kim Vellore
07-07-2012, 2:00 AM
I had the exact problem some time back and landed up making a bunch of files for a range of thickness and based on the thickness of the selected acrylic sheet I used to select the appropriate file.
Glen has suggested some good ideas, I'll try them next time.
kim

Doug Griffith
07-07-2012, 12:08 PM
I used to make hundreds of an acrylic stand and had the same problem. I ended up taking Dee's approach and using a nail file. One of those diamond coated metal ones. My trick was to put the kerf angle to good use.

First I designed the stand so tabs entered from the top side of the cut.

I sized the slots on the tight side toward the bottom of the cut.

Thin spots in the sheet more or less slipped in.

Thick spots in the sheet were filed.

I kept a few variations of the cut file and used the closest one after measuring the sheet.

Richard Rumancik
07-08-2012, 12:43 PM
Glen presented a detailed analysis of the problem and some ways in which it could be resolved. If I understand correctly, his preferred method is to achieve a snug fit and rely on compliance (flex, plasticity) of the material if there is interference. For some materials like ABS this may work fine, but for acrylic (which is hard and brittle) I have some concerns with anything more than a slip-fit as you may have trouble with cracking. And a stressed joint with adhesive is worse yet.

Also, laser cutting has tolerance on the cutting process not just on the sheet tolerance. You will have a difficult time getting the parts that would make a perfect slip-fit every time. So for acrylic, I think that you should expect to file the slots if you don't want loose parts or stressed parts. This labor needs to be factored in, and it might not be practicable for low-end products.

The bumps on the ends of the slots Glen shows could be quickly touched with a file if needed to achieve a slip fit. This would avoid stressing the corners of the slot.

Dee and Doug mentioned filing the long sides of the slot. This can be done, but if you add one or two bumps to the long sides of the slot it will make filing easier. You don't have to file a nice, clean uniform surface, (or get into the tight corners, or risk cutting into the ends of the slot, or make lots of acrylic dust that wants to stick to everything) - you just have to touch the tops of the bump(s) with a file and with a couple of strokes you have removed enough material to make the tab fit the slot. To ensure that it stays in place, a drop of solvent cement on each bump could be added.

Glen Monaghan
07-08-2012, 5:58 PM
Glen presented a detailed analysis of the problem and some ways in which it could be resolved. If I understand correctly, his preferred method is to achieve a snug fit and rely on compliance (flex, plasticity) of the material if there is interference. For some materials like ABS this may work fine, but for acrylic (which is hard and brittle) I have some concerns with anything more than a slip-fit as you may have trouble with cracking. And a stressed joint with adhesive is worse yet.

I'd have to re-read what I wrote, but I thought I said you have to take material hardness and brittleness into account with this method. Softer and more flexible materials allow you to use more lax tolerances (basically equating to having bigger bumps to account for any variances and to get snug fits) while harder/more brittle materials require getting the dimensions closer to correct and using smaller bumps so as to avoid cracking the slots open or having to hand work (e.g., file or sand) slots and tabs during assembly. I've used this with a number of acrylic projects and you can get sturdy, friction fit results without cracking, not a problem but you do have to dial in your dimensions during initial design/prototyping.


Also, laser cutting has tolerance on the cutting process not just on the sheet tolerance. You will have a difficult time getting the parts that would make a perfect slip-fit every time. So for acrylic, I think that you should expect to file the slots if you don't want loose parts or stressed parts. This labor needs to be factored in, and it might not be practicable for low-end products.

Not sure what cutting process tolerance you mean, unless you are referring to the kerf width, but yes, you have to take that into account for the harder/more brittle materials. The kerf width can vary based on material and processing parameters, but is pretty consistent for a given set of conditions, such as 1/4 inch white acrylic cut at 10S 100P. Taking that into account is part of the "dial in dimensions during design/prototyping" I mentioned.


The bumps on the ends of the slots Glen shows could be quickly touched with a file if needed to achieve a slip fit. This would avoid stressing the corners of the slot.

True that, assuming you have a file with teeth or grit on the thin edge, but I've not had to worry about reworking the bumps after a couple of tests and tweaks during prototyping. You might find "sizable" variations at a micrometer level but acrylic bump sizes along with slot and tab lengths are consistent enough between runs to be considered constant in my experience.


Dee and Doug mentioned filing the long sides of the slot. This can be done, but if you add one or two bumps to the long sides of the slot it will make filing easier. You don't have to file a nice, clean uniform surface, (or get into the tight corners, or risk cutting into the ends of the slot, or make lots of acrylic dust that wants to stick to everything) - you just have to touch the tops of the bump(s) with a file and with a couple of strokes you have removed enough material to make the tab fit the slot. To ensure that it stays in place, a drop of solvent cement on each bump could be added.

A few concerns with that approach. First, putting bumps on the sides of acrylic slots requires making the bumps bigger than if done on the ends of slots, in order to account for variations in material thickness. Second, it virtually guarantees having to hand work every slot except for those receiving the thinnest of tabs. Third, you can't build in any ramping effect (other than that caused by the fact that the top of the cut is usually slightly wider than the bottom of the cut). Last, I think the fact that the slot side isn't straight makes the gaps between the slot and tab more noticeable (gaps at the ends are much less visible because they are shorter and often under/behind the tabbed piece. But, some imperfections that drive me nuts like that don't bother a lot of people, so take that last concern with appropriate grains of salt.

-Glen

Glen Monaghan
07-11-2012, 11:23 PM
I was searching for how others have dealt with thickness variations and came across this write-up on using nodes with acrylic:

http://support.ponoko.com/entries/498833-get-the-best-results-from-interlocking-acrylic-designs

Also some good ideas about how to use the nodes for wood and just general inspiration:

http://support.ponoko.com/entries/498853-get-the-best-results-from-interlocking-wood-designs

-Glen

Greg Facer
07-12-2012, 9:14 AM
Yeah, I noticed that too. Designing stuff for a really small router bit makes a better laser design than just using the abilities of the laser to make sharp corners. I know because I prototypes a few joint ideas on the router first, and took out those features to test on the laser. Turned out the router version worked better!

Richard Rumancik
07-12-2012, 10:54 AM
Glen, I see that the second link shows slots with bumps ("nodes") on the long edges - when I suggested using bumps on the long edges you did not seem to be enthusiastic about it because of appearance.

I agree that all of these added features make the assembly less aesthetically pleasing but I never really associated tab and slot construction with decorative pieces. I suppose there are some cases where it can look fine. I have seen some Christmas ornaments and sculptures that look attractive. A tab or two on an assembly is fine, but once you have many tabs and slots I think it is more suited for mechanical assembly (Makerbots, robotics, mechanical gadgets, toys, etc). You will get better assembly accuracy, self-aligning (to a point), and stronger joints than line bonding. In cases of mechanical assemblies functionality and tolerances are more important than appearance.

(I wish the author did not call the bumps "nodes" - somehow I think he made this name up. I think they should be called "nubs".)

In the other link he uses stress-relief holes on the corners. I have a concern with this. The author says it helped a lot to reduce corner fracture, and I have to believe him, but in my experience there is a downside to the small corner holes on acrylic. When you try to cut a small radius or hole with the laser it can sometimes put a lot more heat into the corner than it would on a straight edge. Maybe some other lasers have better motion-control and ramping on corners than mine, but if you generate a large heat-affected zone it can nullify the effect of the stress-relief. The hole on the corners would have to be quite large to avoid this, which would make it impractical. Catch-22, I suppose.

(The stresses in the acrylic can be demonstrated visually by looking at the laser-cut part sandwiched between two crossed polarizing filters. The magnitude of the stress is not easily quantifiable but gives an idea of how far out the HAZ goes. An old pair of Polaroid sunglasses could be taken apart to try this.)

Glen Monaghan
07-12-2012, 11:21 AM
Glen, I see that the second link shows slots with bumps ("nodes") on the long edges - when I suggested using bumps on the long edges you did not seem to be enthusiastic about it because of appearance.

And I also said I was the sort to get picky and caught up in little details that others often don't fuss over... My wife just rolled her eyes when I said one of these things looked ugly-it was made from clear acrylic and had several layers and facets, which she wanted to have assembled without glue. In her opinion, all the clear, sparkly edges distracted the eye and totally hid the small curving gaps at the tabs/slots but those gaps just glared at me. Maybe it's just a matter of knowing they are there coupled with a bit of being an ARSS (anal retention syndrome sufferer, at least I think that's what it means when she tells me "Don't be an arss" ;^)


I agree that all of these added features make the assembly less aesthetically pleasing but I never really associated tab and slot construction with decorative pieces. I suppose there are some cases where it can look fine. I have seen some Christmas ornaments and sculptures that look attractive. A tab or two on an assembly is fine, but once you have many tabs and slots I think it is more suited for mechanical assembly (Makerbots, robotics, mechanical gadgets, toys, etc). You will get better assembly accuracy, self-aligning (to a point), and stronger joints than line bonding. In cases of mechanical assemblies functionality and tolerances are more important than appearance.

Often times, tab and slot does convey a cruder connotation but we've seen some very, very nicely made, good looking, decorative items made with tab and slot on this list. Largely a matter of the developer's artistic skill and attention to detail, IMO.


(I wish the author did not call the bumps "nodes" - somehow I think he made this name up. I think they should be called "nubs".)

In the other link he uses stress-relief holes on the corners. I have a concern with this. The author says it helped a lot to reduce corner fracture, and I have to believe him, but in my experience there is a downside to the small corner holes on acrylic. When you try to cut a small radius or hole with the laser it can sometimes put a lot more heat into the corner than it would on a straight edge. Maybe some other lasers have better motion-control and ramping on corners than mine, but if you generate a large heat-affected zone it can nullify the effect of the stress-relief. The hole on the corners would have to be quite large to avoid this, which would make it impractical. Catch-22, I suppose.

(The stresses in the acrylic can be demonstrated visually by looking at the laser-cut part sandwiched between two crossed polarizing filters. The magnitude of the stress is not easily quantifiable but gives an idea of how far out the HAZ goes. An old pair of Polaroid sunglasses could be taken apart to try this.)

Greg Facer
07-13-2012, 10:31 AM
In the other link he uses stress-relief holes on the corners. I have a concern with this. The author says it helped a lot to reduce corner fracture, and I have to believe him, but in my experience there is a downside to the small corner holes on acrylic. When you try to cut a small radius or hole with the laser it can sometimes put a lot more heat into the corner than it would on a straight edge. Maybe some other lasers have better motion-control and ramping on corners than mine, but if you generate a large heat-affected zone it can nullify the effect of the stress-relief. The hole on the corners would have to be quite large to avoid this, which would make it impractical. Catch-22, I suppose.

(The stresses in the acrylic can be demonstrated visually by looking at the laser-cut part sandwiched between two crossed polarizing filters. The magnitude of the stress is not easily quantifiable but gives an idea of how far out the HAZ goes. An old pair of Polaroid sunglasses could be taken apart to try this.)

From my limited test work (still waiting on my replacement tube #2 from RECI), the heat stress is less of an issue than the focusing of the shear type stress. My brother, running the other laser we bought, had cut pieces where the laser cut 90-95% through the material, but when trying to break them out, the "crack"that is the 95% cut started kept going, right through the neighboring 100% thickness area. So in the right circumstances, acrylic really wants to break. The heat stress is less about mechanical issues like that, more problematic for chemical exposure. We've also noticed that heat stress from the laser is less of a problem a few days later. Edges that can't be glued day 1 can be glued day 2, this in extruded acrylic.

Riki Potter
07-15-2012, 10:56 PM
This seems to be an intriguing topic of debate. I'm going to dabble with Glen idea of adding bumps at the ends, I've made alot of slots in acrylic that hold firmly and don't break anything so it shouldn't be to tricky to get a perfect fit every time when not relying on the thickness.
The idea of bumps on the long side is good to, theoretically would give a more solid hold but I think there would be more breakage problems due to the varying thickness.

Here's a picture of the bowl that prompted me to post the topic. I can assure you that tabs/slots don't always look bad, and we sell these flat-pack so the customer puts them together or it can be posted easily.
236970236969

Richard Rumancik
07-16-2012, 8:59 AM
Riki - the bowls do look attractive - the tabs and slots don't detract at all in this design because the items are made from opaque materials. That makes a huge diffference. When working with clear acrylic, each tab and slot is visible through the material from all directions and it adds a lot of feature lines. Before long the item is so busy with tabs and slots you lose sight of the underlying form. So yes, I would say that in this case the use of tabs and slots works well for you.

Riki Potter
07-16-2012, 5:30 PM
Thanks Richard, I see what you mean about clear acrylic. Even so some designs would work with clear tabs, I'd say the trick is to get creative to use a few tabs as possible and hide/made them look good.
Although at the end of the day not all, probably even most designs wont look good with tabs. On that note what other methods would you use to put it together? Is gluing straight sides the best option?