Just wondering which method do YOU use?

[element]

Just out of curiosity on how others do their work, when modeling a cast part - say an engine block, do you start with a big chunk and remove material or do you build a piece at a time. Same with a plastic molded part - start with a big block and cut away ?

just wondering..

TIA

S

 

[Israr]

"start with a big block and cut away" is not in general a good technique. But it depends on your experience and the type of dimensions etc.


Israr

 

[design-engine]

There is no rule of thumb for or plan of attack.

 

[ankarl]

As design-engine said, no rule of thumb but I do it like this for a casting:


Modeling the core and outer geometry as different parts, then use assembly cut-out to hollow the outside part (with reference to the core part).


To get the machining model I use the merge-feature in the assembly.


Building them "piece-by-pice" rather than carving. (almost always start with curves and surfaces)
Edited by: ankarl

 

[Jacoolo]

with big models as block engine is - as ankarl said, with that difference I used to use copy geom to cut geom inside the part not with assembly func

with small housing I prefer one model for all with the sequence of features: first external geom then internal, it allows you to always suppress all inner feats with no problem and in addition you are sure that external shapes are not related to internal

dratfs, rounds in the end

machined part is created by Copy geom of casting part so You can have control of coloor for machined surf

 

[pjw]

I always usea merge feature for machined castings. Copy geomtry does not work all that well for updating if the original is not in session. Merged feature parts always update (provided 'retrieve_merge_ref_parts' is set to yes in Config.pro. You're better off not starting with a big block of material. Just add what you need feature by feature whether removing or adding material. Don't worry about how many features you have in your model tree - it's not as important as many make out. What is important is good modelling techniques with good use of sketcher constraints, sketch, feature, part and assembly level relations. Try and ensure that most of your drafts and rounds are created towards the end of your model tree. This should hopefully go towards ensuring a more stable model. The mark of good modelling is the ability to make dimensional changes to a model and regenerate without any failures.

 

[element]

Thank you all for the responses.

A lot of my work these days tends to be fixing other peoples work - so I find a variety of modeling habits. I tend not to get a lot of opportunity for creative thinking on my own or extra time to explore within Pro/E even thought I have over 12 years on Pro/E (yes every day).

Even though the hack and slash (big block and cut away) method may not be the preferred I do find in some instances that this can make a better part from a manufacturing point of view if it is done as if you actually were making the part on a mill/lathe - I find that in a lot of cases designers, even those will a lot of experience tend to forget about the ability to make the part in a cost effective manor - nice shapes etc that require multiple setups etc.

Having said this I am going to look at my own modeling habits compared to the suggestions given and see how I can improve - think in a different manor perhapses. I find that it can be very easy to create bad habits when confronted by employers who will reluctantly provide training (if at all) and time lines that almost demand the quick and dirty method.

 

[wsylvester]

this can make a better part from a manufacturing point of view if it is done as if you actually were making the part on a mill/lathe

...... I love to hear that ...manufacturability..

 

[dr_gallup]

this can make a better part from a manufacturing point of view if it is done as if you actually were making the part on a mill/lathe

...... I love to hear that ...manufacturability..

I always do that for machined parts. I start with a piece of bar stock and then every feature is a cut that removes material. However, for cast & molded parts I usually do the reverse. Each feature is a protrusion that adds material. Now if I have a part that is machined from a casting then the first solid feature is a merge of the casting and all subsequent features are cuts.

 

[prohammy]

QUOTE=wsylvester]


this can make a better part from a manufacturing point of view if it is done as if you actually were making the part on a mill/lathe

...... I love to hear that ...manufacturability..

[/QUOTE]


Its funny to see someone being congradulated for making a part manufacturable.....last time I looked that was my basic function as an engineer

.


$0.02 worth coming up


I agree with DrG's approach for castings/molded parts but where I find I differ slightly is where you have a hollowed/shelled casting (which for some reason seems to be the only type I have worked on in the last few months), the further up the model tree the shell is the better in the long term. Too many features before the shell and the number of redefine/regen issues increases quite significantly. After the shell, I tend to find that there is now a fairly even distribution of both cuts and protrusions.


And just to throw fuel on the fire..... when it comes to manufacturability, I find that this is where the Sheetmetal side of ProE can let you down quite a lot. Corner deformation during bending is not what it should be....


Kev

 

[magneplanar]

Prohammy,

here is a crazy thought...avoid the shell

surfaces are more stable...or was it datum planes...hehehe

cheers,

M

 

[prohammy]

I'd love to use surfaces here, but they tend to confuse the hell out of too many of the other engineers that work here (one of the drawbacks of contracting is that you have to fit in)


Kev

 

[Jacoolo]

I feel almost like at home reading prohammy - working among group of engineers(ca 40) which each of them tend to have its own modeling habit - that is crazy.

The neat thing is that each model has to fit in many assemblies, and that is the chalange to convince everybody to do everything right, sometimes not in intuitive but Pro\e way. Add to this big rotation with employment - many out, many new ones - and it is almost imposible to put the average level of skill high.

just my two cents...

 

[magneplanar]

prohammy,

sounds like an opportunity to show them your skill and make a training job out of it?

muadib3d,

have you tried using pro-e on a computer set for a language you do not understand, that will make you pull some hair out...hahaha

 

[Jacoolo]

I faced it once or twice with german seat - not quite "ease of use" situation ))

in genral Pro\E I used and use is in english so it is in not my mother language

but in fact I can understand what Pro\e tries to tell me

I bet magneplanar You have a lot to write down in your CV with such adventure you maentioned))

 

[magneplanar]

muadib3d,

you mean you get paid to do this?

I am just a 3d novice, much to learn...

meetings are fun when they switch to the native language. It is amazing how much a person can tune out if they really have to..." what do you mean you will not work saturday and sunday so you can go to the beach"

cheers,

M

 

[Jacoolo]

well I am paid for playing with Pro\e since 2002. First year it was paying just for pushing my skill higher and higher, next year I was ready to work for the money I was paid, in next years however I knew more than I was paid)))

So that is how it is in life)

 

[verge]

I used to work for a prototyping company, and we used CAD models to generate SLAs, then used the SLAs to make joint boxes for molds, etc. We often wished that design engineers would think more about how they put their models together. The veteran engineers spent a lot of time communicating the possible manufacturing issues to the designers in order to convince them to make a few changes. The idea of building the model based on manufacturing techniques that will be used to make the final product is really what we were trying to convince people of then. Had they done it in the first place, they would have saved themselves a lot of work.


Also, it's great practice for any engineer to learn everything one can - it can't do anything but help you in the end.


And I agree about Pro-E's sheetmetal application. We've been trying to tweak our use of it here for months - it's difficult to actually reproduce what's happening in the shop. The technique always winds up slightly different than what's used in the shop in order to produce an accurate drawing.In other words, if we do it a specific wrong way, they'll do it the right way.

Fun!

 

[wsylvester]

Its funny to see someone being congradulated for making a part manufacturable





Guess it's because I am on the receiving end and I have to render in material. Over the years I have had tocut awaytons of material due to not seeing that manufacturing side of things. That make it out of one part mindset when two would have reduced cost and not changed the function at all. Or that one little detail that doesn't mean beans is sticking up .03 past a standard blank size, forcing me to have to order the next thickness up then having to cut away x number of cu. in. to get there. It starts adding up fast on the larger parts.


I know it sounds like a whine, but it's not. We are not all Nasa, some of us have to turn a profit. but then I'm back in a union, so why should I care

 

[prohammy]

I know it sounds like a whine, but it's not. We are not all Nasa, some of us have to turn a profit. but then I'm back in a union, so why should I care

wsylvester,


Just to pick up on a few things. I wasn't trying to be smart in any way. I spend my life looking at the manufacturability of my designs. There's a simple set of 'rules' I try touse when I am looking at a design


1. Less parts = less cost (This is NOT the be all and end all of a design)


2. More functionality per part = less cost (Combining parts OR combining function, either way suits me)


3. Nesting components = less cost (More components that are the same sheet size the better)


4. Justified Component Function = less cost (I always challange a designs initial spec, I always make people convince me that they need what they think they need. As a contract design engineer this is much easier. I NEVER accept 'Because that's the way we always do it' without a good reason.


5. Design MUST suit manufacturers process = less cost (This is your point wy and it is well made)


The above is not a definitive list, each design brings it's own challanges, but they are the basic way I think of my work as an engineer


Kev


PS I know that this has been a bit of a rant so sorry