Broken glass build stage

I have recently had a number of glass build stages crack when printing ABS. This is strange because I have printed hundreds of parts on this printer without mishap of this type and now every new glass build stage that I put on for ABS cracks.

What I am looking for is any new thoughts of what could be causing this - at the moment I am much less interested in suggestions of how to cure it.



The printer is a Cartesian type loosely based on the early Prusa Mendel pattern. The build stage consists of an FR4 heater with a double Hilbert curve pattern etched on one side. There is no heat spreader other than the un-etched copper layer on the reverse side. The stage is supported on three sprung adjusters feeding the vertical loads as close as practical th the bearings of the Y stage. There are also three posts sliding in PTFE bushes to take the X and Y loads to the stage.



The build stage is either plain 3mm float glass or 3mm mirror glass cut 200mm by 213mm, the edges are beveled and polished. I check each plate both hot and cold with a steel straight edge in X, Y and both diagonals and at the extremes and middle of the Y travel - I have not seen an error of flatness greater than 0.05mm nor any change along the Y travel.

Thermally, the performance of a build stage without a heat spreader is not ideal, but is should be (and has been) adequate. Measurement of surface temperature when the heater is set to 110 C in the middle shows that better than 90 C is maintained up to 20mm from the edge and 35mm from the corners. Obviously this will be worse with the masking effect of a part being printed - however there has been no signs of this causing trouble in the past.

The cracks appear to be related to a diffuse stress of some sort as they are sinuous rather than straight or slightly arced as they would be if they had a well defined origin. In each case there have been multiple cracks, some of which do not extend right to an edge. The crack below in on mirror glass with Kapton tape



The first suspicion was that this was related to Wolfbite as I have recently had chunks of glass pulled from the build stage but discontinuing the Wolfbite changed nothing. Similarly, a suspicion that upgrading two 25mm fans which may have blown air onto the bottom of the build stage may have bought about the cracking proved unlikely after thay were shrouded and the insulation under the heater was improved. Possible mechanical loading from a twist in the Y rails has been checked and found unlikely.

Any thoughts on anything I may have missed?

Mike

Probably a combination of thermal and mechanical stress. I'd recommend replacing the glass with borosilicate, since float glass in the first place seems rather unreliable.

As I mentioned, I have used float glass for very many prints without problem. If I over-engineered at every problem my printer would be built like the great pyramid. Good engineering consists of doing the most with the least.

Mike

IMO heating a plain glass bed to ABS temperatures without using a heat spreader is asking for trouble. Either add a heat spreader, or use borosilicate glass.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

This still does not answer the question "What has changed?" Plain glass can usually take all kinds of abuse within limits and finding exactly what has caused it to reach that limit will allow float glass costing next to nothing to be used instead of borosilicate costing about £11.50

I am not cheap but I don't believe in throwing extra mass and cost at a problem simply because I am unable to see what has gone wrong.

Mike

Ive been using plain float glass directly on the heat bed at 110-115c temps for about 2 years and never broke a glass. Im using the same two pieces I bought years ago.
Im guessing you might be mechanically stressing the glass.
I hold mine on with just two bull clips and ive never had issues. I used to use 4 clips but besides being more of a pain to change sheets it also causes the glass to flex too much using the 4 bolt level method.
I wish this heatbed came with a third hole in the center so I can do 3 point leveling. Trying to level a 4 point system SUCKS.

I wish thermal cameras were cheaper. I'd love to see a thermal image of your pcb as it heats up to see if there are any hotspots forming. Trying a heat spreader would give indirect evidence for uneven heating as the culprit, and even thin aluminium will make a significant difference. But it looks like you'd have to mask the pcb tracks first. Not sure what would have changed to make the problem worse than it was before, unless a section of the pcb has failed, forming a cold patch, or been partially damaged, forming a hot patch.

Just as I haven't broken a glass and have as many as four on the go at once - but now I have broken six without any obvious explanation why. I am fairly sure that it is not mechanical as the heat bed is on a 3 point mounting and the compliance of the bed itself should take up any stress.
Incidentally, leveling a good three point system which has leveling points set in about 1/3 of the way from two sides and one end is a piece of cake. Simply zero the Z and lift to 2mm, move the stage and extruder above each leveling point in turn and adjust using a 2mm drill shank to just clear in turn . If the glass is flat it will now be leveled.

Mike

Thanks JamesK. I previously mapped my build stage with a Raytek IR thermometer and will try it again. I have only checked a couple of points on the bed since getting these problems - a full mapping needs going round 32 points at least 4 times and jotting down the temperature.
You are right, a local hot or cold spot could do it, possibly caused by the sort of damage that you mentioned, or possibly by the heater getting distorted and not fitting sufficiently closely to the glass.

Mike

What about ambient temperature? If you're printing in 20C cooler ambient that would be a significant difference in heat stress, although that sounds pretty extreme.

Although it has been coldish lately, I don't think that it has been below about 12C. I have already checked for cold breezes and all is calm and still. Floyd mentioned the bulldog clips though and that made me realize that I have started using smaller ones than I had used as the larger ones sometimes hit the frame when zeroing. The new clips are definitely stiffer than the old ones - I doubt that this is the problem, but any straw in a storm.

Mike

That's an interesting idea. I use supaclips because of their low profile, but the contact area is pretty small.

[www.amazon.ca]

Might be the clips! I use fairly small ones, they are 3/4" wide and I have no issues.

Eureka!!! I have found it.
I just did a full mapping of the temperatures of a glass build stage at only 90C. Previously, with the heater set to 110 degrees I had only a 20C drop towards the edge. Now, at 90C the temperature is better than 75C except at the right hand side where it drops to 44C !!!!!!!

The culprit is two 25mm fans below the right hand side of the heater, these cool the Ramps/Pololu boards below. When one went faulty I replaced both with Sunon Maglev fans sucking the air from the case - they had previously blown down but I found bits of plastic and even metal which had been blown onto the PCBs.

As this was my first suspect I put shrouds over the fans and a layer of metalised bubble film under the heater - pictures of the shrouded fans and the type of bubble film below.



Checking the air movement with an incense stick there are vortices coming from the sides of the shrouds and playing under the heater. the fact that these would have moved back and forth with Y axis movement would have moved the cold spot about - possibly aggravated by the Maglev fans just having too much mojo.

In mitigation of my failure to see the obvious, where the cracks extended to an edge it was the edge furthest from the fans - the area directly above the fans was pretty much free of cracks. Even the hottest center of the glass plate was where cracks ended, not where they started.

The cure should be to simply re-invert the fans and use a screen above them to sop bits getting onto the PCBs

Mike

Hope this solved it for you!