Yes, the model i posted is listed as a "solid" in the properties display. There are a couple different fillets that MoI does... Edge and surface.. The surface one works best when done on a solid object though can create a fillet with seperated surfaces more robustley, but those will also require a lower level of work to get the result into a solid.. It's a good idea to work with solids as that will then have well defined surfaces and structure.

To get your model into a solid form for filleting I first rebuilt the arcs to use in the sweep as I had mentioned before. After I completed the sweep I seperated everything into seperate surfaces and ran a trim on the surfaces of the sweep object and the hub object, using the "keep" option to keep the exterior surfaces I wanted to keep. Then I ran join on all the surfaces and created a solid form that I posted here.

The fillet of that edge will only go so far before it folds over onto itself.. If you put a .25 fillet on that edge, you will see the results on that little trailing edge patch that is created... Note the difference in the 2 verticle patches that run up to it... This is whats causeing the issue for the fillet corner patch area. If it was a sharp, it will be able to make a more well defined patch, but at some value, you would have to look at it and ask "What would I expect to see put there???" A proper patch structure cant really twist in and around on itself to satisfy the area...

"""" I noticed your model was all one piece and it was able to perform a fillet. Do you have to join them somehow? Being new at this is difficult.""""

It's not really going to be anything you find in the manual for doing what we are talking about here, but those 3 videos Michael has there, cover things pretty well. Even Rhino wont be having specific direction on working in the manner you are describing. Even the higher end stuff may have a need for understanding whats happening under the hood to have success!

Maybe Danny, who is a user here, can find the time to look at your file with NX, a very expensive modeler using one of the more powerful modeling kernels, and comment on the type of results it produces with simple "boolean and fillet" with no need to jump through hoops...

Bueller??? Bueller????

As an example, I created your model as 2 solids "pre-boolean" and sent it to a program that uses the parasolids modeling kernel. Powershape.. Importing the parts in it first starts me through a process of setting up "tolerances" that I want it to use on the part.. Something like a 6 step process.. Then I booleaned the 2 parts together and it started an automated wizard for setting up the tolerances for gap allowance/disallowance and stitching of the surface to ensure a water tight solid.. I then ran a fillet on the edges we are discussing here.. It ran the same size as the MoI file I presented and also created the exact same twisted patch with the larger radius!!! (Than really fits with that geometry)





The red is the underside of the corner patch rolling over onto itself and folding upwards..

At this point, there will be some complicated "Help me fix this area" tools, asking questions that I need to know the answer to, to try and create something there..The help me tools may not exist for the issue and then I will need to know what to do. It may be a week long process to finish just this one component of the model. Then, the very next issue starts all over again... Maybe a months work or more???

Anyway.

@ Bur: Have you the free Version of PowerSHAPE?

Thanks burrman,

I will keep at it. I'll try to do what you did. Did you have to scale the model at all. Or did you just increase the ter. At this point, I'm fine with increasing the ter. It could be that it was increased by the manufacturing department and I just didn't know about it. Everyone keeps focusing on that point, rather than the bigger issue of just getting the model built in MoI. If the ter has to go up then it has to go up, I don't really care at this point. All I have been saying is that is literally the definition from NACA. It is not zero as people have been trying to say. I already attached the document I referenced to write my code (ter = .00021 * chord for the NACA 65A009). You can not bring a zero ter down to a hub and have a working propeller, it will crack instantly under extreme centrifugal forces. Usually the blades are blended from an airfoil to a circle for a variable pitch blade. I'm trying to do a fixed pitch blade so it will blend to something but what I'm not entirely sure (however, burrman is nailing it, exactly what I'm going for). Again, I just want something that is reasonable that either a guy with a grinder made or came off a cnc machine and then was cleaned up by a guy with a grinder. If the naca def is too small, which it may be, then it will obviously have to go up some. But it can't go up that much or the chord will shrink and change the performance. So its still going to have to be a rather small ter blending in to the hub. Again, the idea is that the model would go on for further analysis such as fea. So a realistic model of the blend is my goal.

I've literally never used MoI before, so your patience is appreciated. I spent all day yesterday with it and found it a lot nicer than Rhino, which I have used occasionally. I never could blend the blade into the hub in rhino. It sounds like once I learn what burrman is doing in MoI it probably will explain my failure in rhino. But from what I can tell here MoI works a lot nicer, so I would like to figure it out. Probably all I need to do is learn how to build the airfoil profile better so it sweeps right and then it will work better. Then of course learn how to join it all together to blend it. All of which burrman is trying to explain to me. I'll keep reading, watching tutorials and practicing because it seems like its worth learning and using MoI over Rhino.

The only things I can say from a new users perspective is that the pan and rotate buttons in screen aren't what I would expect. I would think you click the button and then the main action would be those. Not click and hold and move on them. Also when an op fails no output is given. It would be nice if it said the op failed like rhino. Other than that, I think its pretty darn awesome compared to Rhino.

Anthony

p.s. burrman please don't spend two months or more on this. you've done so much already. you have shown moi can do what i wan't and this is just a newbs inability to operate it properly. i won't be here much of the day, so I'll get back at it tonight and tomorrow.

""""" Have you the free Version of PowerSHAPE?"""""""

Hi Frenchy,
Yes.. It's nice as it allows the use of the entire tool suite. If I needed to do work in it and actually save the files, they have an option where you can save a file for $99.00.

Hi prop,
I didnt change anything in the file you posted (Actually, it didnt have the center hub in it, so i copied it from one of the other files and had to scale it to fit.. Maybe you should check it's dims) The only thing I did was rebuild the ler and ter arcs so they were a single segment instead of a broken arc. Same dims though.

We understand that there is a ter radius defined there.. The problem is to model such a miniscule number (miniscule in relation to the rest of the models numbers) is going to present you with very low level problems that need to be addressed. Michael has already stated that MoI "Could" do it, but it is not setup as the tool to deal with what you will be facing with this situation.. I could do the fillet in Rhino too.. His suggestion was, if you really have the need to put that tiny little value into the mix, you may need a more specialized software to accomplish the model.. Or, essentially he will be building it for you over the next month. Every next move will require low level understanding and manipulation (Not in the manuals).

As you point out, many of those blade features are something done by a human, not a machine. I pointed out some info on the naca specs where they actually change the numbers to zero for computational work. If you have a need to model it specifically as-is, for FEA, then it will not be something you wip up in a day or 2.

"""""""""burrman please don't spend two months or more on this. you've done so much already. you have shown moi can do what i wan't and this is just a newbs inability to operate it properly.""""""

Well, not really. It's more like you will run into this with "Any" of the software you attempt this with. The understanding of whats being said about the difference in values with relation to each other, is the key.

"""""""The only things I can say from a new users perspective is that the pan and rotate buttons in screen aren't what I would expect. I would think you click the button and then the main action would be those. Not click and hold and move on them.""""""""

A right clcik/drag in the viewport will pan the 2d ones and rotate the 3d view. Middle click will pan the 3d view..Shift riight clcik will pan the 3d viewport and cntl right clcik will zoom all of them. The viewport buttons are adjustable in the options dialogue. You can make them sensative or not (Like for using them when zoomed way in on something very small.) Their rates are adjustable.. I think they were also designed with a tablet user in mind.. People who use pens to work the computer.

hi,

I played around some in moi and rhino and even rebuilding the arcs and making a solid I still had issues. I thought of a different geometry that should fix everything. when I get it done I'll post it. ill try taking off the ter and with the different hub geo I have in mind it should work out.

Michael,

Here is the model you requested regarding trim not working. It is the same model as before. I just saved it at the point of the trim. The work around I have been doing is just deleting the circle and adding an arc. But if you try to trim the circle, using the two points that lie on the circle, nothing happens. At least for me.

Thanks,

Anthony

Edit; Attachment has been removed to save space on the server.

Hi Anthony, thanks for posting the circle trimming file.

The problem there is again related to the small numeric scale of the objects - your circle there has a radius of 0.0018 and that is too close to the default tolerance level of 0.001 units.

You can solve this by going to Options > General and switching your units to millimeters instead of meters. After you do that your circle will have a radius of 1.8 units instead of such a small numeric value, and then the trim will work fine after that, see the attached example file.

Like I mentioned before, you'll likely run into a variety of problems trying to work with objects that have such a small numeric size - try working with your objects at a larger scale (but don't go overboard into the millions of units though! - think more along the lines of 10 units or so) and things should generally work better.

I'll see about tuning up the Trim command for v3 to use the relative tolerance mechanism which should then make your original example work. The relative tolerance system uses a tolerance that is calculated as a fraction of the object's bounding size so it adapts itself to smaller objects and that works better in situations like this.

- Michael

Hi Michael,

Yes I am starting to really understand your .001 tolerance based on the bounding box. I tried all sorts of ter variations. A point like some suggested, a line as others suggested. Even larger radii. I found, like you said originally, that I would spin my wheels based on the tolerance issue. So I drew a bounding box for the airfoil and worked with that. It went a lot better. All functions worked. I did still have to change the mesh angle from 10 to 1 degree so that it looked smooth on my laptop. But it didn't seem to slow it down any. I created new geometry that may help you or whoever is interested experiment. It is setup to easily apply real life fillets at the base. The blade would need the airfoil ground out as well. So the model is much like you would see in a forging. Prior to being machined. Given the issues we spoke of, this is about as far as I can go with it. But I'm happy with this representation. Its good enough for now. I really enjoyed working with MoI over Rhino, PowerShape, and Autodesk 123D. The forum support was fantastic as well. I have never been a fan of CAD, I am primarily a FEA guy. I have found CAD work to be absolutely no fun. That is until I discovered MoI. I really enjoy firing it up and using it. In fact the more I use it the more a feeling of dread comes over me when I go back to Rhino. MoI is so easy to use. I love the fact that there aren't a lot of different ways to do the same thing. One way for one thing is a great idea to me. The UI is hands down the best I have ever seen. Icons that you can read and know what they mean. I think if you implement the adaptive tolerancing in the future it will really make MoI stand out even more than it already does. I'm going to continue to learn MoI as it definitely seems worthwhile.

One thing I learned through this trial was that Rhino wasn't doing as good as I thought. While I could jack up the tolerance (1e-6 meters) so that I could generate the blade. It still would not make solids from those surfaces and it would not fillet anything. I never realized it was having a problem with the geometry because I was able to work with it. I would much rather have what MoI was doing. Point out the problem right away. Rather than make me think things are ok, when in fact they are not.

The process right now would be to use PROP_DESIGN to generate the blade geo (airfoil profile and rails), model it as much as possible in MoI, and take it to SolidWorks to finish it out and do FEA. I never mentioned this as it was a little off topic, but it may answer a lot of the why am I doing this questions. The geo you see is the ideal shape. Under load it will deform a lot. So I have to generate new rails and iterate to determine the manufactured shape. So you have to go through this process a lot. Therefore ease of use is imperative. There is a lot of iteration required to determine what the real manufactured shape needs to be in order to yield the required shape for aerodynamic performance. Not to mention having to resize everything so it can withstand the extreme stresses. So this is not a one time through CAD and your done deal. For any given model, you will go through this process many many times over.

Sincerely,

Anthony Falzone

Hi Anthony,
I would love to see a screenshot of your models tet or hex meshed and in the FEA program when you get to that stage... It is a very interesting area for me..

Thanks.

Hi Anthony, I'm glad that you've been able to make some progress and that you're liking using MoI!

MoI and Rhino are kind of focused on different goals and so one is not necessarily a complete replacement for the other... For MoI as you've seen the focus is on ease of use and having a fluid workflow and UI. Rhino is more focused on having a large quantity of various kinds of tools, more like a swiss army knife or something like that. Also Rhino is designed to be similar to how AutoCAD functions, which is great if you have a lot of AutoCAD experience but otherwise not so much.

Anyway because of their different goals and priorities it can be useful to use MoI and Rhino in combination with one another - this is also facilitated by the use of the same 3DM file format and you can also use Copy/Paste (ctrl+c/ctrl+v) to move objects back and forth between MoI and Rhino in case you did not find that yet.

re: relative tolerancing - I've been able to adjust Trim for MoI v3 to use the relative tolerance mechanism so that your circle trimming example will work at its small numeric size in MoI v3.


And yes once you've spent some time in MoI's UI you can get pretty hooked on it! :)

- Michael

Sure thing burrman,

There is a free fea program called Z88, that's the best of the free stuff. Caculix is a more capable free FEA program but not easy to use. I think there is hope for Z88 once it develops a little more. You need to be able to do a prestressed modal analysis for propellers. As well as linear static analysis. Z88 only does linear static stress analysis right now. At present, I am trying to come up with a reasonable process flow for anyone using PROP_DESIGN. I have been retired for awhile now, so I no longer have access to all the expensive toys (eh um, I mean serious engineering tools). I do know from my research that SolidWorks Simulation is the lowest cost and most capable program for this application. If you have a seat of that software, I too would love to see some downstream analysis. Currently I'm limited to free or open source alternatives. So I'm keeping and eye on Z88. Scan & Solve for Rhino is interesting but limited to linear static stress of one part. Also, they use a mesh free solution, which should interest you. So no super huge models to run. You don't mesh it at all. Supposedly it compares well to your typical FEA program. But they are a long way from having prestress modal analysis. If you haven't heard of the Intel MIC architecture check it out. That will really make FEA fly. We already have the capability for a lot of memory in a pc which is good to keep the solution in core. If it has to go out of core we now have SSDs as well. So you can build a great FEA machine for 5 to 10k. The price seems to always stay the same but every couple of years the performance goes up. So you can solve larger and larger models or the same model faster.

Hi Michael,

That was fast on the trim fix. I was wondering, and this may be asking for too much. But what would it take to be able to sweep an airfoil and have all the ops work for something of those dimensions. What was breaking down for me if you go past the bounding box was any sort of airfoil shape within that box. It didn't seem to matter if you took out the ter or tried any kind of deviation. It just was plain too small regardless. Then after the somewhat malformed sweep when you tried to boolean it to the hub it would fail. And of course filleting. So I guess what I'm saying is if one op can have adaptive tolerancing can all the ops have it. See, give an inch want a mile.

Anthony

Hi Anthony - the difficulty of relative tolerancing with the sweep is that your case needs to be relative not to the bounding box around the entire curve being used (as was the case with Trim) but instead around the smallest sub-segment within the shape.

That has some potential to possibly make an excessively tight tolerance in the case where someone inadvertently has some tiny micro segment within a profile that they did not really even know about.

I'll take a look at it, but it's a somewhat more complex case with more potential side effects than what I did for Trim.

- Michael

Here's props file back as a solid with .0001 meter G2 fillets at the hub connection and the blade tips... I also changed the hubs cone bottom to be a conic surface with an RHO value of .9999 (Since he likes the little numbers)

I have another file where I made the hub front a G2 blend also, and cut the blade section out to rotate seperatly to try and animate for fun..

I couldnt mesh it. Anything comming even relatively close was an issue for my system.

cool video, how did you make that?

Thats just a turntable capture in 3d-coat... I hadnt cut the blade section from the body yet so you can see the G2 nose cone turning too.. Right now, I'm doing some renders in Carrara trying to get some blur correct for the blade spinning, but it's proving more time consuming than I wanted, because i meshed the blade section very dense and it's dragging things down... I may have to go to sleep and remesh the blades tommorow to allow the render testing to get it to look right.. :o

Hi,

For anyone interested here is an updated model. I was working with the fillets trying to find something both manufacturable and that meets the requirements. This is a close first pass. It leaves room for a blend between the hub and the airfoil. The airfoil isn't ground out as this brings up the tolerance issues talked about. But its a good start and that's all I really was looking for at this point. In this model I increased the size of the forging, changed the hub geometry and put in some decent sized fillets. Everything in MoI works so long as you don't try to mess with the airfoil.

Anthony

Edit; File has been removed to save server space.

Hi Arthur, so I've been looking more deeply into why your original sweep example (The A400M Propeller Example.3dm file that you attached previously) doesn't work right, and I've found a couple of problem areas. The good news is that there is a way to work around them.

The first problem is that the sweep goes through a routine in the geometry library that decides that any edge smaller than 0.001 units is degenerate and that edge just gets left out of the trimming boundary.

So to avoid that, increase the size of your object until the smallest segment in it is at least 0.001 units or larger in length, probably best to have a bit of cushion on that so maybe target something like 0.01 units. This part can be done easily to your example file by just switching your unit system from meters to millimeters (as long as you have not turned off the option to scale objects when the unit system changes which is enabled by default).

So that's the first step - switch to units = millimeters.

Then the next thing that goes wrong is related to having not just one tiny segment but actually having 2 tiny segments right next to each other like this:



Instead of having little arc segments one after the other like that, you've got to delete those 2 little pieces and instead put in 1 single arc piece in there instead (something that Burr tried earlier also), like this:



Having 2 little tiny segments adjacent to one another causes a problem with the joining stage - sweep basically builds one surface for each segment of the profile being swept and then does a joining process to glue those surfaces together into a connected piece. But with those 2 little segments side by side, they end up actually being within the join tolerance of each other and the joiner goes ahead and glues them to each other making a mess in this case. When you have only one little edge in there that won't happen. So getting just one arc piece instead of 2 arc pieces like in one of Burr's previous tries is another key thing.

To replace the arcs, use Edit > Separate to break the curves apart into their segments, delete the 2 little arcs, and use Draw curve > Arc > 3 pts to put in a new larger arc there, then join all those segments together into a closed curve.


I've attached an updated version of your file which should now sweep file - so again the steps to alter your file are - switch units to millimeters, and replace the 2 little side-by-side arc segments at the top and bottom with just 1 arc in each of those spots.


So Burr you were definitely on the right track! :)


- Michael