3D Printer's MAX FLOW RATE TESTING - PRINT vs WEIGHT

[@CNCKitchen]

Interesting results yet from my experience the results don't differ as much. I noticed in your shots that the part cooling fan was deactivated for the blobs but active when printing your part. Did you check if this might be the reason for the mismatch?

[@PrintingPerspective]

I did not just because the real print in vase mode dropped more or less the same as the blob weight percentage-wise, you know. I think the extruder can play a major role here and filament slip-through pushing gears can be a significant factor for the lower extruded blob mass. I remember using Titan vs BMG extruder and the difference was huge with this blob weight method.

[@Ender_Wiggin]

@@PrintingPerspective While I agree that back pressure could be a contributor to lower max flow percentages the results presented are still to variable to make a solid conclusion. I think having more tests and less assumption about what is going is needed to find the true flow rate. For example with fan cooling you introduce more heat loss to the heater block therefore lowering print temperature by a small percentage. What this shows is that the true flow rate will depend a lot on the geometry you are printing and the specific factors of your printer (hot end, heater power, exact nozzle size, material shrink rate). In the end we can only test so much. Finding a more representative value for flow rate is valuable. Edit: The comment about thermal expansion/contraction is very much spot on. If you care about percies wall sizes this method may work better but is not a general solution. I still feel that weight of model and blob are still a more consistent measure as a general solution.

[@noodl3d]

This methodoligy is flawed. line width is not taking into account the the thermal expansion/contraction factor of the filament. If this is taken into account then the results would be closer if not spot on. So in theory the contracted line that is less than the expected width would be denser than a line ovet the expected width. Leading to the same weight....

[@noktrnl123]

This is getting extremely interesting.

[@Vez3D]

very nice video man. Another possible way of measuring flow is with filament lenght extruded versus requested. I did some test with that technique back then because I didnt have a good enough precise scale like Stefan's method. And It was very interesting to see the results. Thanks for sharing this one and the deep dive.. I love that :) Keep it up

[@Avets610]

Very interesting results. Also thankyou for condensing them in a single graph.

[@PrintingPerspective]

Glad you liked it!

[@skilledcamman4752]

All your tests are always so detailed. Nice Work!

[@hjaltesrensen8214]

I really enjoy how you are able to explain and how so clearly your point in your videos, you don't just use the practises of the printing community and take it at face value, but in some cases show the real problem or like here contradicts a common calibration method. So cool to see, and happy to see my name at the end of you video xD Keep up the good work

[@PrintingPerspective]

Thanks for the support :) Comments like these give a lot of additional motivation. It is always very nice to hear from people who understands, sees, and enjoys the goals of the content I try to provide. Thank you!

[@Redo3D]

The way I benchmark my hotend is I take a miligram scale, I cut exactly 200mm of 1.75mm filament and weigh it, its usually around 0.580g. I then go to my printer interface (Fluidd) and Extrude 200mm of plastic at a set speed. I weigh it, and I continue to either increase temperature or slow down extrusion speed untill the extruded filament matches the weight of the 200mm of raw filament I cut. Once I find out this exact balance I save the settings and that is what I use for alll of my printers in my print farm. I get perfect extrusion on all of my machines this way.

[@LWUndsoo]

Pretty neat idea, i need to do that too

[@PrintingPerspective]

Your method seems like a blob weighing method just manual.

[@Redo3D]

@@PrintingPerspective Yep it is! But its pretty easy method to do when you have a lot of machines in the print farm. Once I am happy with a setting, I perform this method on all of my machines and just tweak the rotation distance of the extruder for each printer to match the weight.

[@OmikronPsy]

This method is so clever and practical, especially when many printers use the same filament! You could even use the underextrusion factor and increase flow by this value if speed is more important than quality, like for infill.

[@jim5148]

Thank you for taking the time to do the research and make the video!

[@xilw3r]

Nice tests, I like that this channel is looking into lesser talked about topics which quincidentally bother me too. Although I'm more interested in how to get closed loop volumetric extrusion, among other things

[@PrintingPerspective]

Thanks, I talk about things that bother me too, haha ;D

[@stefanguiton]

Great work @diyperspective! On another note, I was wondering if you could explore closed loop vs open loop stepper motors in a future video. I have seen a few claims that they increase torque and decrease stepper motor heat. Would be interesting to see this effect on accelerations and print quality.

[@PrintingPerspective]

Thanks! There are a lot of topic that interests me but to be honest they are just not viable to make videos about as they are extremely niche. As a full-time small creator, I have to think of ones that will interest the more general audience and bring valuable info to a more hardcore audience. Otherwise, it won't be a sustainable job, unfortunately.

[@802Garage]

Very cool! I recently did some extensive flow rate testing with the blob method as well. Since doing the testing on various upgrades, I have definitely found the printing results don't always like un perfectly with my data. I suppose that is to be expected when changing so many variables like print speed, layer width and height, and infill amount. I will say the blob testing has given me great general guidelines for maximum printing speed though. Part of what I thought about most when testing is what you alluded to. The faster you print, the less hot the filament coming out of the nozzle actually is relative to the overall temperature reading. This will affect print properties and it's part of why if you are printing near the limits of flow, it may be beneficial to turn the heat up slightly not only to prevent under extrusion, but also to maintain layer adhesion and uniform heating of the extruded filament. When printing a very large print at high speeds for example, near my max flow rate, turning up the temperature just 5 degrees improved the quality of the surface and possibly avoided the line stacking you were talking about. That was with PETG. Note that performing this same testing on PETG for example may produce very different results. It was very surprising to me how differently flow rate testing went with PETG compared to PLA. Anyways, good stuff, look forward to more tests!

[@PrintingPerspective]

Thanks. Now I am very interested in how layer adhesion comes into play with all this. Max flow with PETG from my experience can depend a lot more on the filament itself. I have some that are impossible to print above 6mm^3/s and others print fine above 10mm3/s on the same printer. :D

[@802Garage]

@@PrintingPerspective Yep I saw a lot of variance as well. My Creality transparent PETG flows way easier than my Longer black PETG+ for example. PLA is more consistent, but at the same time the higher temp and generally higher quality PLA tends to have a lower max flow rate while the less durable stuff flows fast. I was really surprised my Creality carbon fiber PLA was actually the highest flow rate, because I like its properties overall and it doesn't print super low temp either.

[@danielrioux54]

This video is excellent and very thoughtfully produced. It delves into an important distinction that is often overlooked: the difference between the maximum flow a printer can output, the expected flow rate, and the flow rate at which the part will be dimensionally accurate and mechanically sound. The explanations are clear and detailed, making it easy to understand the complexities involved. The visual aids and examples provided effectively illustrate the points being made, helping viewers grasp the nuances of printer performance. One important recommendation for future videos: please slow down when you speak. You clearly have an in-depth understanding of the subject, but for the audience, it can be challenging to keep up with your rapid pace. SLOW DOWN, SLOW DOWN, SLOW DOWN, PLEASE. Overall, this video is a valuable resource for anyone interested in understanding printer flow rates and their impact on 3D printing performance. The Stefan’s Blobs test, for instance, demonstrates the maximum amount of plastic extruded for a requested flow but does not indicate the part's mechanical performance. Just because a printer can extrude a maximum amount of plastic does not mean the resulting part will be high-performing. Additionally, this test is static with limited pull on the material exiting the nozzle, which does not mimic real printing conditions. On the other hand, the material deposition test offers a more realistic representation of the actual material flow rate. The graph presented in the video clearly illustrates this point. Mechanical performance, as the video rightly points out, depends on numerous factors related to the material and its properties when melted, deposited, cooled, and various printing settings like retraction and temperature. These factors are too complex to be linked solely to the flow rate. The main goal in high-velocity printing is to support short processing times. Evaluating the flow rate as the ultimate indicator is a mistake because this is a multi-dimensional issue. To address this, I suggest using a process called "Design for Experiment" to identify key factors that ensure parts are produced quickly, efficiently, and retain their mechanical properties and appearance.

[@T.J]

You can slow down the video by clicking the little gear icon in the lower right. Personally I appreciate the speed at which he talks and understand it all but given his accent and pace I totally understand it could be difficult to understand.

[@ZhechenZhu]

Great content. Keep it up!

[@PrintingPerspective]

Appreciate it!

[@m00se-b7t]

My flow rate testing using the vase mode print method also using an hgx lite and chc pro matches your values quite closely.

[@PrintingPerspective]

Thanks for the info :)

[@zbyskie]

2%-5% could well be just the plastic shrinking more due to the higher stresses put on it from printing faster. I especially noticed that i had to slow down my ABS profile quite a bit for this exact reason.

[@PrintingPerspective]

The A5M Pro printer results were quite good even at high speeds because of the long meltzone. I think measuring extrusion width at different flow rates is great as that also gives you data how different filaments perform during various flows. Those couple percents might not seem much but that also will affect how consistent your layers will look. So it's also good for finding the filamets that flow more evenly at different temperatures.

[@ThePhilbox]

Very useful info!

[@PrintingPerspective]

Thanks!

[@MrRitzyVlogs]

Should we expect a percent error on a linear dimension to match with weight error (i.e. a volume error)?

[@PrintingPerspective]

Weight error mainly depends on the print test, and measuring width of the extrusion you have to practice a bit to get error free measurements. That's why I redesigned my print model as it was hard to measure and the older version introduced other variables. So with more practice, better model, and measuring 4 sides of the extrusion I feel it gives very accurate readings to judge how each filament performs.

[@ivanabratte6647]

hi man great video about this comparisons, do you think that in differences with the blob test and the print test, the consumption current of the motors X and Y may cause a decrease in the current in the Extruder motor, causing that the flow decrease slightly? Maybe is a chance of this happen, you can try this using two separate motherboards, one for the axis movement and other for the extruder motor with separate power supply.

[@PrintingPerspective]

I doubt it is the case as the weight of printed parts was very similar to the weight of the blobs. Weight will never take into account the fluid dynamics of the molten plastic when extruded at different temps due to faster printing speed. At this point, I think measuring extrusion width is a significantly better way because it is a realistic printing scenario and it gives way more valuable data on how plastics actually flow.

[@alainthire]

Gold content!

[@uhu4677]

Very interesting findings. However I think the weight method is a more accurate and relevant than the thickness of a one-wall test-object.

[@PrintingPerspective]

If you say so, you should also explain why you think so, you know.

[@VorpalForceField]

Nice work..! Thank You for sharing.. Cheers :)

[@vahvelpoiss]

I just use the Orca slicer max flow test and it works rather well for real world testing

[@PrintingPerspective]

Yeah, it is pretty good to determine the maximum flow. The only thing is that you can't measure the true extrusion width and how much different filaments drop off. And that is a very interesting data.

[@MisterkeTube]

I feel this is mixing up max. flow rate measurement with correct wall thickness tuning. CNC Kitchen's blobs are great to figure out the max. flow-rate and even give information on how the flow-rate is affected the more filament you try to push through per second. That is a very valuable characteristic of your extruder-hotend-filament combination. However, it is not the only parameter determining wall-thickness. As the video points out, even if filament still flows good enough to come out, it might not spread out as easily as it does when heated longer (so with lower flow) and that will affect wall thickness, but that effect itself (so not just the wall-thickness, but the amount that thickness deviates at higher flow) will also depend on your selected layer height and line-width, so it's much harder to characterise than a single max. flow-rate value. In short: the blobs give you valuable information on the absolute maximum flow for extruder-hotend-filament, but you might still have to lower your filament throughput (or tune the slicer's line widths or max. speeds for certain line types) if things like exact wall-thickness are important for your print.

[@PrintingPerspective]

What is the flow rate? It is how well extruded filament flows and determining when it no longer flows to be printed. At higher temperatures, plastics flow better and at lower worse. With filament spending more time in the hotend, it gets hotter, and printing at higher flow rates it spends less. The decrease in extrusion width literally shows a real-world scenario of comparable states between flow rates, how well the filament was heated, and how well it flowed. That affects how consistently the layers will look and how well they bond. You can get so much valuable data from that. And that is my goal here. But if you think other methods suit better for you, I respect and have no problem with that.

[@MrEksol]

А как сделать тест с печатью кубика?

[@noodl3d]

You are not taking into account the thermal expansion of each of the materials.....different materials/even different colors within the same material can expand or contract differently. This factor has a big influence on extrusion width.

[@PrintingPerspective]

I tested multiple different PLA filaments with different colors and the trend was more or less the same between the two testing methods. I personally want the filament that reacts to different extruding temperatures as little as possible. That gives more consistently looking layers. So measuring the extrusion width at different flow rates gives the most meaningful data. I don't get the point you trying to make here.

[@Geoff_W]

Umm... I really suggest you get a different scale. I had that same scale and it drifts a LOT, you need to recalibrate in between each measurement.

[@PrintingPerspective]

I think you have a bad one. I never encountered that, there is a tiny deviation but mine work quite consistently. I even checked now weighing objects back and forth.

[@Geoff_W]

@@PrintingPerspective I went through 4 replacements of theirs before I gave up and returned it, but yeah if your re-validating with the weight in between measurements it's probably fine...but your data just isn't anywhere close to the results I see in my own testing. One of us is doing something wrong 🤣

[@PrintingPerspective]

@Geoff_W you talking about the blob weight method, right? What extruder and hotend combo you have?

[@Geoff_W]

@@PrintingPerspective Yes, blob vs pattern method I ALWAYS see the blob method as lower than the pattern. AnkerMake M5c is one I have that you also have, for example.

[@PrintingPerspective]

@Geoff_W That is so strange, as I never saw the blob method drop more percentage-wise compared to the extrusion width on all my tested printers. Might be the filament you are testing with or maybe the extruder not having enough tension. It is really surprising to hear that.

[@sabahoudini]

Why not weigh the printed parts, just don't add any brims etc? Should be much more accurate. Maybe print the first layer first and weigh that and subtract it from the rest of the measurements because it will be printed at a lower speed. Measuring width seems not very accurate.

[@McFlyOrPie]

AAAAHH! Cliffhanger!