MikeTrice

I think your dimensions are a bit out. LNER cross section below:

You may find that just reslicing it solves the issue but the other aspects are worth checking as they don't take long compared to a failed print. Good luck and we hope to hear of any progress.

Not quite what I said. You need to open the sliced file rather than the source stl. Chitubox, Photon Workshop and UVTools can all be used to do this. THis will show whether the artifacts are caused by the slicer. It might be worth turning off any anti-aliasing/blur settings if used as some slicers are currently having trouble with them. Given the stl slices and prints correctly on other printers why would the problem be in the CAD? Duncan mentions in his query the fact he already printed at 45 degrees. 0.051 layer height sounds odd and might suggest using 0.05. Personally I often print with 0.04mm layer height. It is also worth checking to see if there is a later version of the firmware available for the printer as some of the Anycubic ones still have bugs. It would not harm to try this it there is room on the build plate.

Very strange. No obvious explanation but in the absence of any other suggestions I wonder if there is a slicing error. Are you able to open the sliced file to check the individual layers for signs of the issue?

www.3smr.co.uk ?

From an old Peco catalogue:

I had a look at that and the basic figures are nicely detailed though lost at 4mm scale. To be honest for $68 it did not give me anything more than I could do with MakeHuman.

Are you referring to exporting from Blender as an STL? All objects in Blender will export as a single STL which can then be imported into Chitubox as a single entity. It does not matter whether they are on a layer, the layer is visible or not, or the object is visible or not, they will ALL export. If you want separate objects you will have to copy the blend file and delete all but the ones you want and then export that.

For this sort of work I usually turn to using Inkscape to try and scale things. Measurement on Google Earth suggest a width of 20'0" and I have used this as a basis. Attached PDF is reproduced to 4mm scale so if you print it with no scaling you should be able to guess widths from it. romaldkirk.pdf

It is different from more "normal" resins in that you need to give it a post printing heat treatment. As its name suggests it is tougher than normal resins and can be machined. It has gained popularity for use in 3D Printed Wheels.

If anyone is after it eSun Hard Tough resin is back in stock at Amazon uk.

Those are they.

Cornice plates over the doors?

Brian, photo 2 of John's post suggest there are recesses as they can be seen.

This might help. In an attempt to make things more obvious I have created a locked layer which has been filled with blue. I am now drawing on a new layer above it. Start off drawing your basic rectangles for the wall and openings. At this stage they are all filled with white (which is immaterial): Select all your rectangles: Now select Path->Exclude: You now have a wall with the windows cut out hence the blue of the background shows through. Import and place your Scalescenes texture behind the wall: Select your wall and background texture and select Object->Clip->Set: The background texture now fills the wall but not the windows:

That is the method used by the LNER.

Hard to tell without seeing how you supported it. Beware that an open wagon will by its nature create areas of suction that can rip things off of supports. You might need to add a couple of vent holes.

If you have not already done it fit a screen protector. I am running mine at 80%. I will at some point do an Anti-aliasing test which will involve reducing the power further and increasing the exposure times to greater than 3 secs.

Good question that I cannot remember (it has been a while). I found a format that gave 1920x1080 23 fps and saved as mp4. When you upload to Youtube I think it reprocesses it.

I use OpenShot Video Editor.

Jan, suction is not quite as you state. Imagine a hollow tube printed vertically. If the tube is printed flat on the plate it effectively forms a cup when pressed against the FEP forming an enclosed hollow space. When the piece is pulled away from the FEP a suction effect will occur with the trapped space which increases the effort needed to separate from the FEP. This effect can impact the print causing distortions and even splits. If you raise the tube away from the plate and use supports the unvented void does not occur. Unvented voids aside there is a peel force exerted during printing separating the last printed layer from the FEP. Angling the work piece might be to reduce the cross section of the print that needs peeling off or to reduce layer visibility. So in the example in the previous post a thin rectangle can be printed vertically on the plate without a suction effect. Hope that makes sense.

I cannot see any issues with the sliced .photons file.

Rob Apologies, I had not realised that the roof was solid. Yes you need to reduce the exposures when using clear resin, however my experience of printing clear is limited. Those distortions do look very odd and I cannot offer an explanation. If it was me trying to print this I would hollow the model and raise it above the build plate and rotate is some 70-80 degrees. I am assuming here that you have a side that can be placed nearest the build plate and have supports going to it. There will be some distortion where the feeds attach but hopefully they can be sanded smooth. You would then need some supports to keep the model steady during printing. If it helps feel free to PM me the stl and I will do a sample chitubox file and return.

Following on from Mikes previous answer it could be related to suction as the canopy is flat on the plate. Agree regarding raising from the plate (which will solve any suction issues) and angling. Looking good though.

Interesting post on the Facebook Anycubic Mono X Owner's Group regarding antialiasing. They state that there is a current firmware issue that fails to apply anti aliasing at layer exposure times less than 3s (the Firmware cannot process the AA fast enough at those speeds). Short term you need to increase exposure times above 3s and reduce UV power. New Firmware is being worked on.