What Fillament, wall thickness and Fill Percentage for Patterns

[Jasonb]

Can I ask what Fillament people are using for their patterns as well as the wall thickness and percentage of fill.

These are for patterns that will be sent to a foundry so may not be handled quite as gently as ones cast at home. Sand will be Airset. Castings upto about 2.5kg (5lbs) in iron.

Had a few failures of the first patterns off the Bambu P1S particularly when the 0.2mm nozzle was used. I'm not doing the printing so it's a bit third hand but will try and get the settings used.

 

[creast]

I am not sure if this helps but for sand casting where the patterns could be used again I generally print in PLA with a 0.4mm nozzle and set wall thicknesses to 1.2 mm and 20% infill is usually ok.
I also do lost PLA where I reduce the wall to say 0.8mm and infill of 10% but I guess you are not using that technique.

Rich

 

[Jasonb]

Thanks Rich, they will be used as traditional patterns and withdrawn from the sand and get more than one use.

 

[Scott_M]

Hi Jason
The Bambu P1S has an enclosed chamber and can print ABS very well ( I have one too ) The big benefit to using ABS is that it can be vapor smoothed with acetone.
I would suggest using a .4mm nozzle with 3 walls and 15-20% gyroid infill. And a small layer height like .1 or .16mm. This will make the layer lines very fine. And then vapor smooth it.
Vapor smoothing is really easy to do. I use a plastic container with a snap on lid like a big butter tub or Ice Cream tub obviously big enough for your part. I put a few thumb tack on the lid and set that down and put my part on the points of the tacks so it is elevated from the bottom. And take the tub portion and put a paper towel in the bottom and hold it there with a few magnets inside and out. Soak the paper towel with acetone. ( not enough to drip ) and place the tub over the lid with the paper towel above the part. In about 15 minutes it will get very smooth. Sometimes you have to go a bit longer but anything over 30 minutes and the whole part will get soft and may distort. When you think it is done take the top off and leave the part sit in open air on the tacks for at least an hour. It will be tacky and could take finger prints or distort, so leave it sit a while.
Maybe ask for a few small ABS parts to practice on. I have even sanded seams before vapor polishing and it even removes the sanding marks.

Edit: you don't have to snap the lid on, you could just set the tub portion of it over the part on a bench that won't be hurt by acetone fumes. Part must be elevated to get the vapors on the bottom side.

But... I have never done any casting or made any patterns, this is just 3D printing advice on getting strong smooth prints.

Scott

 

[a41capt]

Thanks Rich, they will be used as traditional patterns and withdrawn from the sand and get more than one use.

I haven’t done any casting with my prints, but if longevity is what you’re looking for, I second the opinion regarding a 0.4mm nozzle. I’m also a big fan of PETG as it’s a bit more impact resistant, resists warpage due to heat, and once dialed in prints very well at 0.16 layer height. As far as infill, I would suggest a gyroid pattern at not less than 30% and, as you said some longevity is an issue, at least 4 to 5 top, bottom and outside lines.

Extends print time, but makes for a very durable print. Here is a short video of a 3d printed steam (air?) engine I built, only metal involved are fasteners. Still holding up fine even after an Arizona summer and temps well over 110 Fahrenheit:



John W

 

[Jasonb]

Thanks for those pointers, I'll pass them on.

I think they were going for speed so used a thin wall and not much fill, still waiting for details but this is an image of one of the patterns being printed.

Probably worh printing a few of the same part with different settings and then seeing how they hold up to a bit of rough handling.

If it ends up that they will be getting more use then we will print off an enlarged version and cast an aluminium pattern from that

 

[a41capt]

Thanks for those pointers, I'll pass them on.

I think they were going for speed so used a thin wall and not much fill, still waiting for details but this is an image of one of the patterns being printed.

Probably worh printing a few of the same part with different settings and then seeing how they hold up to a bit of rough handling.

If it ends up that they will be getting more use then we will print off an enlarged version and cast an aluminium pattern from that

Looks like about 10% infill. Even then, if they switch to a gyroid infill it should add a bit of extra strength without much time increase.

No brim shown, so they’re getting good bed adhesion with the filament employed. This cuts down on their post processing time and speeds up delivery of the finished product.

I’d love to see the finished print, and of course a finished casting!

Thanks,
John W

 

[awake]

You mentioned using a .2mm nozzle - this is going to be very, very challenging on any FDM printer. .4mm is about as small as I would want to go; anything smaller, and you are likely to spend more time unclogging the nozzle than printing!

 

[mrehmus]

You mentioned using a .2mm nozzle - this is going to be very, very challenging on any FDM printer. .4mm is about as small as I would want to go; anything smaller, and you are likely to spend more time unclogging the nozzle than printing!

Bambu's 0.2 nozzle seems to work very well. There are some filaments they recommend not using with it but the finished print can be extremely smooth.

 

[Scott_M]

A .2mm nozzle will make corners sharper and provide finer exterior detail but will not do anything for smoothness. Layer height is what makes prints smoother.

Scott

 

[Jasonb]

I believe the 0.2 nozzle will do smaller layers than the 0.4 hence why it was being used to get less obvious steps on the tops of convex surfaces thus reducing the amount of hand work needed on the patterns..

Thanks for the other info it's been taken on board.

 

[awake]

Bambu's 0.2 nozzle seems to work very well. There are some filaments they recommend not using with it but the finished print can be extremely smooth.

Can't argue with success!

 

[Scott_M]

Jason
You are correct. The minimum suggested layer height for a .2mm nozzle is .06mm and for the .4mm nozzle it is .08mm, not a lot, but smaller. But if time is a concern the .4 nozzle will print a specified wall thickness twice as fast.( or 1/2 as many passes.

Fun fact
The Bambu nozzles have a very large flat on the bottom and you can print extra wide line widths. With the .4 nozzle you can print line widths 1mm wide without problems. I am not sure how wide a .2 can do but I would guess it will do .5mm wide. This feature is extremely useful for unsupported overhangs. With a small layer height and a wide line you can almost go horizontal. You have to go slow and use lots of cooling but it works.
And the wide face is also very good at ironing the top surface.

Scott

 

[GreenTwin]

This conversation illustrates the need for accurate record keeping of the settings that one uses for their 3D prints, and also a record of what works for others.

I have unfortunately generally let the Prusa slicer pick default values in the past, and have not kept records for any of my 3D prints, so that is something I really need to start doing.

I did reduce the infill on the last few prints from the default settings, and that saved time and filament.

It is the same way with foundry work; if you don't keep accurate records (or remember) things like optimum aluminum and brass/bronze pour temperatures, exact additive amounts, mixture ratios and set times for bound sand, core gassing time (5 seconds max with CO2), etc. then you tend to make the same mistakes repeatedly.

.

 

[Jasonb]

Thanks for the replies, I've suggested the guy doing the printing asks here himself so he can give the details of what settings are being used etc rather than getting things third hand through me.

Scott, I have suggested they use the 0.4mm combined with the softwares variable layer height option but don't know if they have used that or not. It should help with the near horizontal yet still be faster on the vertical.

it's a bit similar with CNC if you just set the passes to either a horizontal or vertical stepover you will get "contour" type lines and surfaces that are close to perpendicular to the stepover but it can be faster. The best option is to select a path that steps over as set distance across the actual surface then each pass is the same distance from the next no matter the angle of the surface. Maybe when the bring out 4th axis printers they will be able to do the same "scalop" type path either by rotating the print or turning the nozzle so it is at an angle to the surface.

 

[Scott_M]

Maybe when the bring out 4th axis printers they will be able to do the same "scalop" type path either by rotating the print or turning the nozzle so it is at an angle to the surface.

Now that would be COOL !!

Scott

 

[ennech]

I have experience using ABS printed patterns for sand casting. It is true that foundries do tend to mistreat pattens, I find mostly in trying to extract them from the sand, many times by digging in a spike to do it. That's ok with a wooden pattern but ruins a 3D printed one. I use a brass insert particularly in the case of split patterns. This is normally referred to as a "rapping and lifting insert". It is made from a piece of hexagonal brass and has a tapped hole 1/4" BSW through it. It is retained in a suitably shaped hole in the pattern by either Araldite or epoxy putty. I use Milliput Superfine which sticks to ABS without issue and smoothes out easily using water. It's also handy for adding the odd filet to a pattern.

 

[Grumage]

I would like to thank Jason B for opening this thread and to all members that have contributed thus far. I have been involved with model engineering for over 40 years but a complete novice, nay beginner in this field of engineering.
I have made hundreds of patterns over the years, I would now say, the hard way! Father Christmas arrived early on Christmas Eve with a Bambu Labs P1S which has seen a crash course in virtually every aspect of things unknown to me personally. We have already gone through a whole reel of filament with a couple of patterns breaking at the foundry. They mainly use Air set which does have a tendency to grab the pattern and I think we were trying to skimp a little with the pattern making material.
Having followed the advice from several members here the engine bed pattern, shown below has all the latest information built into it. We’re currently using the 0.2 mm nozzle, which has been mentioned as problematic, but we have had no issues with. However perhaps someone could explain what happened to the initial layer where a slump has occurred. It’s no problem as the Cope is moulded on a flat surface.

Cheers Grum.

 

[ddmckee54]

Congratulations on your first 3D printer??? Welcome to the nut-house/rabbit hole? 3D printers should all come with an "Abandon all HOPE ye who push the start button" warning label. I'd have 3 warning labels now.

If you are saying, like I think you are, that Layer 1 when printed is wider than Layer 2 and above - this is commonly referred to as "Elephant's foot". It's caused by your nozzle being closer to the bed on Layer 1 than the controller thinks it is. That extra volume of plastic needs to go somewhere, so it's oozing out to the unconstrained side/sides.

Some slicers have an option called Elephant's Foot Compensation which allows you make Layer 1 smaller in both X and Y, to allow for this oozing. If yours' doesn't have this feature, you need to adjust your Z homing limit switch upwards a little bit. Symplify3D doesn't/didn't have it, Prusa Slicer does, and Slic3r does - I don't know about Cura.

Don

 

[Jasonb]

If you click on the thumbnail twice it zoomes in and looks as though there is a small bead intended to run around the inside approx 5 layers high. Crisp at the end and only a bit of a blob on the longer internal edge so may be something other than nozzle height. Again if it here Nozzel height related I would have thought it would also show on the outer edgeGraham or Alan can you post a screen shot of the part in Alibre?