Looking for suggestions

[harley558]

I had to create this disc design with these depressed areas. I am looking for any methods that would acheive the same result that I did. I feel there is a simplier method out there, I just can't think of it (sort of like can't see forest for the trees)

DISC.zip

[JD Mather]

First thing I noticed is that most of your sketches aren't dimensioned/constrained. Surely you didn't create these in AutoCAD?

Second thing I noticed is that the part is nearly, but not quite, uniform thickness. Looks like logical design intent would be uniform thickness in which case a bit of changes in your features and Shell would simplify by about half.

[harley558]

This was based off of a cad drawing and they wanted to see a 3d model. As far as dimensions there are none to go by at this point, it is still in concept phase.

[JD Mather]

As far as dimensions there are none to go by at this point, it is still in concept phase.

 

As a machinist I am not familiar with working without dimensions. Good luck!

[harley558]

As a machinist I am not familiar with working without dimensions. Good luck!

 

I could supply you with dimensions but for what I am asking they would be irrelevant at this point because we are not sure of the pilot hole size nor the bolt holes until we get confirmation from client to proceed. That is why I said it is still in conceptual stage. What I asked about was how to create the depressed area better. Maybe since SHIFT works with wheels he can answer this question seeing how you are deadset on it being dimensioned when the sketch came from Autocad why dimension it again. At this stage I am not concerned with the constraints, I am just trying to develop this model.

[shift1313]

Hey harley. I would probably approach the model with the shell operation in mind. There is probably a dozen different ways to think about this type of a model. Instead of removing the pockets, you could add the "spokes" back in, still thinking about using the shell operation. The "spoke" method is probably the first route I would go. Also shell the model before you add the bolt holes.

 

Unfortunately I dont have the time at the moment(maybe not tonight) to draw anything up for you but I will as soon as I get the chance. Are you using 2011 now or are you on 2010?

 

I will try to get to it tonight for you but i cant make any promises.

[JD Mather]

There is probably a dozen different ways to think about this type of a model.

 

Here is one of twelve. Yours on left, mine on right.

[shift1313]

harley. Here is a quick attempt at the shell method I was explaining. I wont say its the easiest way and after the first attempt I noticed several areas that could be improved but the general idea still stands.

 

I did make one basic modification to the disc you posted. Instead of the pockets almost going to the outer edge, i made the face of the lip be the edge of the pocket(hope that made sense). So each "spoke" and the lip all have a 35degree draft on them. As i said there could be changes to the approach for this method and remove a few things from the feature tree so don't pay too close of attention to every aspect of it.

 

If i get a chance i will think about another method for you.

Disc2.zip

[harley558]

Thanks JD and Shift for your assistance...Shift I am using 2010 is there any way you could reproduce that in 2010 so I can examine the steps you went through to acheive that.

[JD Mather]

I noticed shift dimensioned and constrained his model.

[shift1313]

Thanks JD and Shift for your assistance...Shift I am using 2010 is there any way you could reproduce that in 2010 so I can examine the steps you went through to acheive that.

 

Sorry harley, i have no way to produce it with 2010 now. I might be able to draw it in an older version of inventor that you could use. If not I will try to take some step by step screen shots to help explain it when i get a chance.

[harley558]

Thanks again, it would be most helpful if you do have a earlier version i.e. 2009 or earlier then I could just use 2010 to open it and review how you did the design.

[shift1313]

Hey harley, i made another file but even zipped its just barely over the size limit. Rather than remove some features can you send me a PM with your email address and i will send it your way. It was made in 2007 format.

[Hopinc]

Hi Guys,

 

I created the model in 2010 using one less step than JD, but I did have problems with the shelling.

Unfortunately I could not quite get the 0.156in thickness without errors. I did get to 0.150in.

 

Out of curiosity Harley 558 is this thickness dimension critical?

 

I copied you model as faithfully as possible - no dimensions. There are of course implied constraints.

 

Regards.

 

Dave

[shift1313]

Hopinc, less steps doesnt always mean a better model. You just want to make sure to eliminate any extra steps you may take. There are also things like Fillets and patterns that are better to do outside of a sketch as a feature, rather than in a sketch. While in your feature tree this may look better as a sketch, in the long run you will want the flexibility of making them using features. Im not saying you did it wrong, just saying dont get hung up on the number of steps you take.

[Hopinc]

Hello Shift1313,

 

I was not suggested that less steps make a better model, I am not in anyway hung up achieving that end. It never enters my mind when I am modelling. But in truth the less steps in a model, the less calculations are required.

 

In practice the reduction of steps in a history tree is aided by experience, but in real life modelling this is seldom a factor of importance. I am sure that like me, if you could redo some of your earlier models (24 years worth in my case) you would create them a lot quicker and slicker.

 

My illustration was only to show that there was another way of modelling the part, albeit with one step less than JD's method. As far as I am concerned the original post asked for a simplification of the part modelling process and I provided evidence that it is possible.

 

..... I feel there is a simplier method out there, I just can't think of it (sort of like can't see forest for the trees)

Regards.

 

Dave

[shift1313]

I agree with you dave and i apologize if it sounded like I was saying your method is wrong in any way, that wasn't my intention. I always approach models from as many angles as possible from the start and like you said it comes from experience. I try to focus on making the models robust and easy to modify/update first and foremost. I just wanted to make sure new users don't get hung up on a simplified feature tree rather than exploring options on how to make the parts. We can sit here and tell people how we feel its best to model a part, but trial and error can go a long way to help understand why

[Hopinc]

Hi Shift1313,

 

We are on the same wavelength. In fact your model is a much better, more manufacturable solution.

 

Sometimes though, I find it is best to show a user that what they want to do is achievable, exactly they way they wanted to do it. Then gently lead them on to a better result - if there is one.

 

Trial and error is definitely a powerful method of learning, and one I wholeheartedly advocate. Unfortunately it also has a downside - frustration.

 

Best wishes to all.

 

Dave