Formlab 3 precision is worst than FDM

Hi everyone
After a long experience on FDM ( ultimaker 3 ), I started using the Formlab3 using Durable Resin, to get more precise engineering mechanical parts.

I am actually disappointed from the precision, which was supposed to be extremely better than fdm, and I got actually bad result compared to my PLA FDM print. The surface quality is much more better, post processing much more easier too, but I got a part of 3.3mm instead of 3.0mm .
This part is supposed to fit into a “female” part, with a 0.2mm playing, and as a conclusion, doesn’t fit at all.

I printed in 0.05mm height layer, durable resin V2 ( see screenshot below ) , whashed 35min , let it dry for 2h, then cure it 60min at 60°.

Actually i have no idea how to improve this print, as preform doesn’t give me so much parameter. Do you have face any similar experience ? How did you solve it ?

Thanks a lot

edit : I also “sand” all the support , so the surface I am measuring is completly flat

Recommended wash time for durable is no more than 20 minutes.
Try to print again and measure before and after washing, see if it’s getting thicker because of wash.

There is also a calibration option in the settings (via printer’s touch screen), you may try it.

Hi @IIHMForm3,

Thanks for reaching out about this - I’m sorry to hear about the poor results you’ve had so far. As @agiorgitis mentioned, it’s certainly possible that the part has swollen due to solvent overexposure. If you are still seeing less than ideal results after adjusting the wash time (see Form Wash/Form Wash L time settings), I would recommend reaching out to our Support Team.

1 Like

Thank you for you answers @agiorgitis and @Jesse_K .

I was wondering, because the poor print part is close the bottom ( according to the angle I gave on preform software , cf printscreen ), is there a chance that this orientation reduce exposition light during curing - because really close to the floor ?

Should I try multiple orientation as a calibration, and choose the best one ?


1 Like

When i said calibration I meant this guide. It’s about printing a part and entering the values in your printer.

You must wash and cure the calibration parts at the same duration/settings as you’re going to do with your other parts.

About your own parts, you can also add one more measurement, after curing.
So: Measure before washing, measure before curing, measure after curing. Then adjust accordingly.

Try moving the part on the build plate, you’ll get different results back compared with forward edge.

Hi @IIHMForm3,

Trying out multiple orientations and multiple placements on the build platform could certainly be worthwhile. The calibration above adjusts the X/Y fit of your parts, and could also yield some improvement based on the orientation of the part.

I would also rotate your parts about the Y & Z axis 45° (as you’ve done in the X axis) so that the pull forces are minimised for each layer.

This has the added benefit that it normalises the aggregate of the layer thickness across parts that have many flat faces with vectors 90° to each other (as your part has).

Hope this is helpful?

1 Like

I’ve been designing and building custom tools using all kinds of 3D printing technologies for more than 12 years (with over 25 year in tool making background) and unless you CNC parts to spec no 3D printing will ever produce super accurate parts. you print a test part, measure all needed Dimms adjust you design and print again, and again if needed. There are some many, many factors that can cause your change from original. Even if you go back to your statement regarding FDM and super calibrate your printer it will still not last as such. In CNC world we have tolerances and compensations for a reason.
good luck.


Thank you for all your replies.
I tried an other print, more or less a TEST print, with differents orientations. I tried to be precise on the process as I used the exact cure and wash time.

As a result of the print, the thickness seems to be really more precise if the piece is be printed vertically. In others words, maybe Z axe is disturbing the accuracy.
As a comment, 3.04 and 3.16 are similar as is depend exactly where the calliper is placed. But still, between 3.21 doesn’t fit in my 3.2 gap and 3.16 does.

What do you think about it ?

I understand your frustration on this but why not just change your CAD design to 2.9mm or 2.95 and try it?

1 Like

Consider minimizing your wash time.
In the lab I manage, wash time (with clean IPA) is never more than 5 minutes.
With some parts, depending on geometry, IPA wet wipes are plenty sufficient.

Also, thank you jantarek for highlighting this for parts needing tight tolerances

1 Like

This has been my experience since using our Form 3 for the last year or so.

It’s not the z axis that is causing the issue. It’s the nature of this type or printing. The closer a face is to being parallel to and facing the build plate the worse the accuracy of the position (and details) of that face will be. Parallel faces that are normal to the build plate exhibit high accuracy. Faces parallel to the build plate and pointing away from the build plate exhibit the highest level of detail achievable.

Are you able to share the model file (and which resin is being used)? I’ll print some off and do some measurements to compare.

Hope this is helpful.

Hello everyone, and thank you for yous answers :slight_smile:

@GGagnon : of course I can change de CAD design, but I have already included “playing” et “tolerances” in my cad file. There is already 0.2mm gap between the parts, so I don’t care really that the piece is 0.1mm thinner or thicker, but I want the piece to fit and turn inside the other. So this is more consistency than accuracy that I am looking for. I the case of the 3.21mm, my problem is that the spacing in which the piece should fit in is also 3.21 ( so really accurate )

@Andrew_W : this is the part :

@RD_lab : thank you for the tips

@jantarek : yes you are right, than only CNC can have precision better than 100 micron

This is the result on the FDM machine. So yes, precision is good because it is a multipler of the layer height ( 0.1mm ), so maybe accuracy is not the right word. But let say that FDM machine doesn’t have to be jalous from Formlab at this point.

Let me reclarify what I meant in my reply, I agree with you on one thing. I wish 3D printing was “plug and play” with tolerances and especially when you are calculating a clearances or fit into your design and even 3D print those objects at the same time and yet - its a no go, or no fit :face_with_diagonal_mouth:. There are so many factors to consider when 3D printing functional parts: type of material, 3D printing technology, Print orientation, supports, postprocessing, etc.
In my case, 3D printing completely changed how I do things in my profession. FDM offers cheap fast prototypes and ready to use parts as well, SLA offers complex geometries and materials I could only dreamed of 10 years ago, I have not use (personally) SLS yet but its coming to my library shortly. Lots of folks here have experience and great knowledge to help with printing solutions so sharing those type of “trouble’s” is great. But I do not agree with statement that FDM is more accurate than SLA, rater you might need to review the technology and material you need for this build or redefine your design if this its the technology and material it needs to be. Anyway that is my 2 cent statement and I hope I didn’t step on anyone’s toes :wink:

Just because everyone is going to place the stl differently on the place, is it possible that you place it in Form and then export/share the form file? So everyone to print in the same orientation, same position, same supports. Then compare.

I printed off 3 of the parts on our Form3 and 1 part on our Ultimaker 5S for comparison of dimensions. I used Grey Pro as we have a lot of this resin.
As an aside the SLA parts took about 2 1/2 hrs and the FDM part was about 1 hr.
Dimension are in the images and seem pretty good, the parts all fitted together easily.

Thanks you @Andrew_W for this time and printing.
The orientation you gave is really interesting.

I printed all my differents pieces in the same vertical position.

Results are pretty good and everything fits in.
I still have a rotation problem, as the round shape of the “female part” is not perfectly round.

I made a video than explain my problem, where the common hole, at the center of the rotation, should be always align - as a shaft hole. Or, in the 4 configs on the video, there is always a problem when the long piece is SLA, especially at the end of the rotation.
Here is the video :

@jantarek : thank you for putting science on the top of the debate ! you are definitely right, with the shape I want and material properties I need, I have actually better result on my Ultimaker, but this not a generic answer for sure.

This is because part of the surface is facing towards the build platform. The nearer a surface is parallel to the build platform the lower the positional accuracy. And all your clearances are consumed by this. That is why in my above posts (and prints) I rotate the parts in all axes.

Also be certain to wash out all uncured resin from narrow grooves before post curing.

1 Like