How To Approach Complex Designs | Walking Through the Design of a 3D Printer Nozzle Vent
ฝัง
- เผยแพร่เมื่อ 28 ธ.ค. 2022
- In this video we are going to tackle a question that came in about surface modeling and forms modeling, but we are going to solve it with solids. Often times planning out a complex design might point towards a solution like Forms but end up being much more difficult to pull off.
In this video I will talk about my design approach when presented with a problem. How I think about the design paths and what some of my steps are.
For this design we are going to look at modeling the internal flow volume or void as a solid body. Essentially making the "fluid path" as a body that can be subtracted from something else. This might be an unconventional approach but often times the issues with shelling or thickening a complex shape will drive you mad!
a360.co/3Q0cdmS - แนวปฏิบัติและการใช้ชีวิต
![ไม่ต้องมีที่ที่ให้ฉันอยู่ แต่ขอแค่มีฉันอยู่ก็พอ (Spaceless) - getsunova [SHORT FILM]](http://i.ytimg.com/vi/MuT0d_cvJkc/mqdefault.jpg)
Happy to learn something unexpected: curvature map analysis was a very helpful trick to identify problematic areas on any operation that uses an offset: Shell, Thicken, Offset, ...
Thanks for adding these gems!
Glad it was helpful!
You designed in a few minutes something that I spent hours and hours trying to design. Thank you for the shortcuts. I never thought to use lofts between predefined profiles, etc. I was busy combining sweeps, etc.
Glad I could help! Sweeps might be a good options as a starting profile but sadly we don't have tangency control on the profile of a sweep in Fusion. So I would use it as the first solid/surface and then maybe loft from there.
This was super helpful. I never would have thought to model the internal shape first but I can see why you do it this way. This has given me a lot to think about. Great video Matt!
Glad it was helpful! If you have any more questions just let me know.
Learned a lot from this. Like the real problem real solution(s) approach to meet the objective.
Great!
Thanks, I learn a lot from your channel. You definitely deserve more subs and I recommend your channel to anyone that asks me about design on fusion.
I appreciate that! Thanks for watching and spreading the word!
First❤ love u videos, some times when I stuck and look some complex stuff from you I get inspired
Love that!
Hi! Thanks for all your videos, they have been a huge help. Around the two minute mark, you touched on how thickening/shelling a form can cause errors. Do you have any content on workarounds for converting complex forms to solid bodies? Thanks!!
Thanks Nathan, it is such a case by case basis unfortunately. The big thing to look at is the curvature map with the Maximum. If your surface has any red areas on it at the thickness value you want then it will likely cause issues. The only ways around that are to work on the form some more(which the took can be on and active in Forms) or to trim and patch up the surface after the form work is done. If you have a specific example that you are stuck on and would like to share I would be happy to work through it to show the approach.
@3:50 you describe why you chose to model the interior volume of the duct. Does that mean you when you are worried about constant shell thickness, you typically chose to model the outside of the part? Maybe you could do a video expanding about when to model in, or model out, dangers, pitfalls, and and traps.
@16:13 I thought you were going to mirror the second duct, but then realized you couldn't because the primary duct wasn't symmetric. You could have done this by forcing symmetry in the drawing and then you'd only have one set of profiles to make pretty. Disclaimer: you're definitely a better modeller then me, and you've tought me quit a bit about Fusion stuff I had no experience with.
Hey There, the choice to model inside vs outside is a decision based on a few things. For this, the inside shape of the duct where the airflow is was really the key design element. The outside shape didn't matter so getting the inside right was key. Second is a thickness thing. Since the inside flow was what I wanted to control if there were thickening issues then fixing them on the outside wouldn't affect the function.
On the symmetry you are correct. This was answering a modeling question someone sent me so the inlets/outlets and general process I worked within those confines. Working with symmetry whenever possible is 100% the way to go!
Great video, Caducation! I've actually been working on a hotend assembly like this and was wondering how I would go about modelling the ducts.
On another note, would you be able to make a tutorial detailing how to design a compression screw for injection molding and/or filament extrusion? They're different from normal screws as the volume/shape and, resultingly, compression factor change dynamically along the length of the screw 🤔
thanks!
on the compression screw, I think the pitch and thread stays the same but the center tapers down. Something like this? www.vestamid.com/en/processing/injection-molding/media/292868-desktop.jpg
@@LearnEverythingAboutDesign Hi Caducation, I believe you're absolutely correct. The slope surfaces (flight thickness) and pitch seem to maintain consistent dimensions throughout the screw; whereas the channel depth decreases towards the valve/exit size (due to the spine gradually increasing in size). Continued...
=====
I realize that I wasn't specific enough earlier as I should have been, but my request more-so lies with a very special compression screw from the ARTME MK1 Filament Extruder project as depicted here (pages 2/15 and 3/15 on the pdf) www.artme-3d.de/wp-content/uploads/2022/08/Mk1-mechanical-parts-22-08-17.pdf
Here, the pitch (from the center-point of each flight) seems to be the only variable that remains consistent. The slope surfaces (flight thickness) and channel depth, however, seem to change dynamically over the length of the screw.
Additionally, these compression screws seem to have a much thinner spine overall.
Any tips on how to tackle something like this? A tutorial, also, would be more than welcome~
Thanks for the link! Yes that looks to have a consistent 15mm pitch but the rest is changing. Let me give it some thought. This is definitely not Fusions strong suit without a variable pitch "coil" that is table driven like some other CAD programs.
Hi, great content, I love how your videos aren't fully scripted, rather you take us with you through the process and arising challenges, I learn a lot! That said I would like to ask a question - I just started learning Fusion recently, I have lots of fun with it, but occasionally there is a problem with curved surfaces. You are the first person I saw using the Curvature Map Analysis and I think it's going to become a very useful tool in my arsenal, but could you maybe do a more beginner friendly introducion to it or is it a tool not meant for begginers? I searched on your channel for videos explaining it but they fly a bit over my head, I can't find explaination how you would approach problem like shown in this video with the curved air ducts using curvature analysis to overcome it rather than finding a different way (or to find out if there even is one). Or did I miss already published video explaining that?
Thank you!
Its a bit of an advanced tool for sure as generally when you need it you are outside the realm of beginner type geometry. I have covered it inside of many videos but specifically in 2 videos focused on the tool.
th-cam.com/video/Z34fcjtpjEY/w-d-xo.html
th-cam.com/video/yWSafAfDjxI/w-d-xo.html
the first one was in the context of freeform modeling for car bodies. The second was looking at surfacing tools and seeing where we might have issues on a thicken. That is really the main use case looking at the thickness value you want and if there are areas of your design that are under that radius value.
I use it on complex shapes to take a look at the flow of edge details. Like on car bodies ill look to see how wide the red lines are. If they should be fairly consistent vs getting wider or narrower (indicating a fading edge). The tools is "live" meaning you can turn it on and keep modeling.
Another tip is that a lot of the complex tools have an analysis tab on them now like surface loft where you can enable it on the preview of your feature. It can be helpful there.
@@LearnEverythingAboutDesign Thank you for the reply :) Yeah, those two videos were the first I watched trying to come up with a solution but maybe I'm not yet at a stage where I can extrapolate ideas that easily. I will get there, though!
Love the content just wish you went with a more printable design considering its a 3d printer fan shroud.
Right well the question came in and the user was asking about process. They were planning to break it up into multiple pieces and glue it together so the initial question was more about workflow than the restriction of the build direction. This could easily be printed with an SLS machine and the lower portion of the vent could be split so the upper half was printed and then the lower half.
Great video but I still have a question. How would you model this if the ducts came out of the fan at an angle to end at the bottom instead of just going straight down?
Depends on how you need to define the shape. Lofting between profiles on angled planes would probably be my first approach.
Interesting why the loft surfs had self intersection issues at 1mm offset. I don't know much about Fusion 360, but if you had the ability to throw a bunch of flowlines in the U or V direction on the surface you could better visualize what's going on. Often the rendered image and even the curvature maps don't really tell the whole story. Lofted surfaces can exhibit all kinds of problems because of the uneven spacing or syncing of the curve parameters involved. Wiggly or uneven flow can result in shape problems
there is a curvature analysis that will toss UV curve lines on and allow for curvature combs to be displayed as well. Fusion doesn't handle these lofts great. If you keep the input profiles the same number of edges it certainly does better. Lofting with a Form body usually produces a better result, but the profiles aren't adhered to like a surface or solid loft and the number of edges that need to be added to say have a slot with two lines and two arcs isn't achievable as a form.
There is no link with the subject of the video but I notice that there is no joint folder, how is this possible, how is your assembly and these sub-assemblies built?
Sorry about that. Check the description now.
The design was sent to me, i didn't build it but the only time a Joints folder exists is if there are actual joints applied. This design contains a sub-assembly of the carriage, a sub assembly with the fans and then one with the mount. Things are in place but nothing is constrained. If you aren't going to have/show motion it isnt required to have joints or even ground the components.
No problem, Thank you for the clarification.
I made this comment because I want to know what is the best approach when you want to do a full top down assembly. There are several possibilities, ground, joints, rigid group, but there is not much about this process in the documentation. "Things are in place but nothing is constrained. If you aren't going to have/show motion it isn't required to have joints or even ground the components." > indeed, if the "component drag" option is disabled
@@mac_ito well Fusion will always ask you to capture the position or revert if you try to have an operation. If I wasn't going to show motion I might consider everything a body BUT as this is likely part of a larger assembly having them as components makes sense.
I do have plans to cover workflows but its a MASSIVE topic and I have yet to figure out the best way to convey it. There are a lot of design decisions that happen along the way. The "rule" was to start with components for every new part, but in practice i vary my approach.
OK thanks for these notes.
I come from solidworks and it is not easy to find an efficient workflow while keeping a lasting mastery of what you are doing. at least that's my feeling.
@@mac_ito For sure. So there are some major differences right. In solidworks if you use the Master Model technique which I used to use, you have multiple solid(or surface) bodies in a single part file. This is essentially how Fusion works if you are just creating bodies. The nuance here is that in Solidworks you could have under defined sketches and have Instant 3D turned on and it could behave a bit like an assembly. But at the base level you can think of Fusion as behaving like a Master Model multi-body part file.
With Fusion you can convert those bodies to components at any time, which is similar to converting a master model in SWx to an assembly. The main difference here is that in SWx that conversion creates a new assembly file while Fusion it just stays in the same file. When you convert a body to a component in Fusion you activate it to continue working(although it isn't strictly needed.....) Each component has its own coordinate system and Fusion keeps track of its location and orientation. thats how joints work by relating those.
If you decide to start your fusion file by creating empty components, activating them and starting the modeling process you are essentially doing a top down assembly like in Solidworks. The difference between this in solidworks and Fusion is that everything is still timeline based. So best practice would be to create components and apply joints early(which you can even do without solid bodies).
This doesn't take into account a few other nuances like direct modeling, base features and turning the timeline off.
My general workflow has been to maintain bodies until i need motion. There are cases where i might start with a component and the main benefit here is that the timeline stays a bit cleaner. When you activate a component you are only seeing the timeline for what was done to that component. You don't really get that when you convert a body to a component.
There are other workflows like having each part external and inserting them into another file. This is called Distributed Design. This is helpful if you have a common part across multiple assemblies. It can maintain the link. You can even edit it directly using Edit in Place which will attach some XRefs to it from whatever you select. In my opinion this gets messy and I avoid modding parts in this manner.
You also have a derived workflow. This is similar to the solidworks master model if you were to take a body in your feature manager , right-click and Insert into Part(or new part i cant remember). Basically you take a body, bodies, components and you can pull them out of a design and into their own design. This is helpful for manufacturing if you want to build jigs/fixtures outside of the main file. It can also be helpful if you are designing an enclosure and work needs to be done on the inside by other users. It prevents you from having to reserve a file and possibly lose work. The flow of data is uni-directional. Changes to the original file flow down to the derive, but not from the derive back up.
The problem with covering all the workflows in a video is that the nuances and decisions evolve and the workflow changes. but what i said above should cover the basics.
From my experience, fillets are great for cnc machining but horrible for a 3D printed part if not very large. You have to think about layer thickness and how many layers it will take to step over and create a clean curve on the fillet. Often, it's better to have sharp corners. How do you get around these printing defects with a design like that on such a small part using a 3d printer.
Print orientation helps. Desktop FDM machines have good curvature in XY so if the required curves can be oriented in that way it helps. Smaller layer lines like a .1mm layer if using an FDM. Or use a resin based process like mSLA.
I agree that adding a chamfer/sharp corner helps for sure. This video was done to help a user trying to design a smooth vent/fan housing flow path.
@@LearnEverythingAboutDesign Those are good things to consider. Thank you.
How about this. After your undersized model is complete (nozzle-1) make a copy (nozzle-2). Lets say you want 2mm walls. So upscale nozzle-2 by (in this case about 110%). Then, using the Move Tool and Section Analysis, position nozzle-1 in the center of nozzle-2 on all 3 axis. Then cut nozzle-1 from nozzle-2. Don’t keep nozzle-1 (the cutting tool). Then make three Offset Planes where the three openings are and adjust the planes 2mm (or what ever is needed) inward. Then cut the three 2mm parts from nozzle-2 and delete them. Do you think this would work?
If i understand correctly yes, but there is likely an easier way. Instead of scaling if the intent is just a thinwall part you can just go to the Surface tab and use Offset to select the faces you want and offset them outward by 2mm. This is probably the easiest way to do it.
@@LearnEverythingAboutDesign But wouldn't this give you a wall with zero thickness? Or can you thicken the offset surface?
Yup from there you just use Create > Thicken
@@LearnEverythingAboutDesign Thanks. Dang Fusion 360, a million ways to do most anything. Have you seen any of the guy "Fusion 360 School"? He really does some weird stuff with F360. My problem is I'm 88 so I sure don't have the time to learn the stuff you guys do. Thanks again.
@@larsord9139 Yeah he does some very cool stuff on that channel! I like how he does case studies for certain geometry. I try to not duplicate his style as I think that works very well for him. A lot of it is tribal knowledge so any way to consume the info is great!
With surface tools it helps to think about the designs one face at a time. It gives you the freedom to create some pretty complex shapes and then worry about joining them together after :)