Flow Calibration with Bambu TPU bonded to plate

I performed a flow calibration for Bambu TPU but the resulting pattern fused to my Textured PEI (Gold) plate.

  1. Suggestions on how to remove it without damaging the plate?
  2. Suggestions on avoiding this in the future?

And yes, I used glue on the plate. (Bambu’s liquid glue stick)

Put in freezer for 5-10 minutes and then try again.


Didn’t you calibrate the filament first ?
The flow test should not cause this unless a normal print would fuse as well.

For ‘fused’ prints there is no fix…
But if just stuck on the bed the above mentioned freezer approach is always good.
You can also try to soak it soapy water as it creeps everywhere.
If all fails try the shock method:
Put a big zip lock bag in the freezer that you filled with a bit of water.
Place on a FLAT surface as that is where the print will go.
Pre-heat your oven to around 80 degrees celsius - then check if it is not already above 100…
Place the print in there for about 30 minutes then quickly onto the FLAT side of you ice block bag.
Wait until you can feel the lower parts of the model getting cold.
Won’t come off? Repeat…

Unless you really scraped the nozzle over the bed and used excessive temperatures the print should pop off eventually, just don’t use brute force…

1 Like

Perhaps I’m missing something: how does one ‘calibrate the filament’ apart from doing a flow calibration?

In my case, a model can generally be coaxed off the plate because its ‘Z’ dimension is wide enough to make the plate separate with flexure, whereas the flow calibration is only a few layers thick so the pliable TPU print just bends with the plate and does not loosen. I attempted separation using a scraper, but it left a thin film of TPU in some places.

Thank you for all the suggestions.

Great question!
I asked myself this one many times in my early days…

Of course the flow rate IS the main factor here.
But the temperature is major player as well.
You can print PLA seemingly fine and with little no visual changes from about 175 to over 230 degrees Celsius.
And IF the speed ALWAYS matches what the filament needs the print will also be structurally fine.
But our machines have to slow down for certain things and not always do all layers take the same time to print.
Finding the RIGHT temperature that neither lets the filament go brittle nor risks loosing the layer adhesion can be vital.
Let me try to explain:

If you print solid infill than the extruded plastic can only go one way.
If you print walls then usually the inner most and outer most wall are the same as this single infill line starting over and over again.
A high nozzle temp allows the plastic to spread out easier and faster - creating a different result compared to printing on the lower temp side of things.
A flow rate calibrated perfectly with a vase mode print usually results in solid infill to build up - resulting in rugged edges towards the wall or if too bad a file like surface.
Do it the other way around and most certainly the vase mode print won’t have the correct wall thickness.
A change in nozzle temp affects this and if it is just right the difference between the two is minimised.

The K-factor affects the flow rate predictions based on the speed and acceleration.
And again: a big difference in print temp can affect the K-factor result.
If the factor is too far off you can get differences in the outer layer, like visible lines of slightly ‘misplaced’ layers or rough bits.
And if way off in the other direction thin walls with just 2 or 3 layers might come out really bed if they have corners…

The max flow rate is a bit of a ‘use if required’ thing…
In a nutshell the max flow rate limits how fast the filament can be extruded.
Is vital is you want the fastest possible print speed but also if you like to combine infill to the max.

All these little things do add up, especially if the model allows for these things to turn out really bad.
Like the nightmare of so many people, supports…

For me and in most cases it is the surface finish and accuracy that matters most.
Thing like the bottom of a box resulting the the wall poking out while the walls above are further is is unacceptable.
Same for a DIN spec nut and bolt not fitting together…
Doing this now for about 20 years and if you ask me than the claim that our modern printers provide perfect prints out of the box is about as true as what a politician tells you these days…

I once removed a bonded TPU phone case from the plate by getting one corner pulled up with pliers then pouring alcohol under the corner. But by bit I worked it off that way.