UM2 Build plate temperature very uneven

[blackomega 2]

As a side line to something else I'm investigating I decided to see just how uniform the build plate heats to. I was surprised by the results.

For reference the build plate temperature setting on the UM2 is set to 105C and a build is in progress.

1. View from the top looking down on to the UM2. You get a sense that the build plate is not being heated uniformly but in this vision mode it is affected a bit by ambient air temperature.

2. Same top down view but this picture is in contrast mode and ambient air temperature noise is much reduced (or almost eliminated). You can see a much greater contrast between the high temperature areas (white/red) to the low (yellow/orange).

3. Same as above but this time with the spot temperature readings. As you can see there is a massive difference of 10-20C between areas of the build plate.

 

Is this to be expected or does it hint at a problem with my build plate heater?

 

[blackomega 2]

As a fun bit, you can see where the UM2 logic board is mounted! lol (pics are taken facing the UM2 from the front). I

 

[gr5 2,167]

Is this to be expected or does it hint at a problem with my build plate heater?

 

That looks reasonable. I've seen much worse when the screw heads stick up and create an air gap between plate and glass.

You will of course get better printing results for ABS if you enclose your chamber. But this is probably good enough for most prints.

Why do you suppose the left side is worse? Maybe inspect the distance between glass and aluminum there with something thin like paper. Were the fans on? I try to print ABS with the fans off or at least on a very low setting.

 

[jonnybischof 60]

I love these FLIR pics

Did you leave the color <-> temperature scale away intentionally? That would've been a nice information...

 

[blackomega 2]

I love these FLIR pics

Did you leave the color <-> temperature scale away intentionally? That would've been a nice information...

 

Nope my FlirOne doesn't seem to want to show the temperature scale otherwise I would have. You're right though it would have been nice.

 

[yellowshark 153]

Depends on how long you waited before taking the pics. If I am printing a big piece I have to wait 10mins+ for the perimeter areas to get to the desired temp after the central area has achieved it.; but in my experience they do get there. No idea though on how using a glass plate, which I do, affects the question.

 

[jonnybischof 60]

Glass plate... Glass plate... Glass plate?

Can an infrared camera even measure the actual glass plate? Shouldn't it just go through the glass and display the temperature of the heater beneath the glass?

Could it not be that the actual surface temperature of the glass is more even - and some 3-5K colder - than what the image actually reads?

20K differences are big though, this will surely not be equalized completely.

I remember playing with a FLIR cam at my previous workplace - when you looked at a window, you could see your heat signature's reflection on the glass, but of course only as a very slight shade. Most of it just went through it.

 

[Dim3nsioneer 557]

Refraction (transmission) is not relevant for thermography. The reflectivity is. You have to take the background into account for infrared images. And - this is important - you have to use the right emissivity (usually called 'epsilon') for the surface if you want to have a quantitatively correct image.

 

[gr5 2,167]

Trnsmission does matter. But not for glass at these temperatures I believe because I think glass isn't transparent at these infrared frequencies. Glass is transparent in the near infrared but hopefully by the time you cool to 100C (from 6000C) it's opaque.

I know for sure glass is opaque at the infrared frequencies when things are at room temperature. I know this because at night I get dew on the glass winshield of my car. If the glass was transparent down to these frequencies the dashboard would be the item getting "radiatively cooled" and the dew would be on the dashboard.

Some materials like quartz are transparent down to these frequencies. The camera used in the above photos obviously doesn't use regular glass!

But, yes, reflectivity of glass *does* matter. But if you look straight on (not at a steep angle) very little light is reflected and mostly you are seeing the "color" of the glass (the temperature).

 

[gr5 2,167]

This subject: emssivity, transitivity, reflectivity - how these 3 values always add up to 1 (or 100%), how emissivity ALWAYS exactly represents both absorbed and emitted light of a given material, how this affects dew, how this affects how things look in infrared cameras, how these properties can confuse infrared cameras, why there is dew on the windshield and why it is on the grass, but not the grass right next to the house - all of these issues are very interesting to me and I have spent 100s of hours contemplating them. If you want a little more detail... ask away for clarification - please try to figure out what part is confusing first.

 

[yellowshark 153]

I finally got around to looking at this again. I switched on my printer and set the bed to 76c. to achieve 65c on the glass plate. Well actually I start with the temp up in the 90s to speed the heat transfer process up but once I get to around 60c I then turn the heater down to 76c. Once the central are was measuring 65c with my ordinary digital thermometer (does it use IR) the perimeter areas were lower, ranging from 5c to 15c lower.

I left the printer for 30 mins and on return all the perimeter areas, bar one, were within 1.5 degrees of 65. The upper rear left of the plate was about 3 degrees short. I took the plate off the bed and placed it on a wooden table and saw exactly the same temps.

 

[axsdenied 0]

As gr5 said, the heads of the countersunk screws holding the clips can be sticking out leaving an air gap between the heater and the glass plate. Check the heads of the screws in your cold corner, that could be the cause of your problem.

Mine were all sticking out and I improved countersinking of the holes. The average temperature of the plate went up by almost 5 degrees C.