Machinists CAD Tool - Looking for ideas

I've been considering building a web based CAD like tool that would make various calculations easier for day to day machinists and engineers. This would be something short of a full CAD application, but would be a time saver for simple calculations.

Ideas:
- given a target hole size and tolerances, calculate the closet matching size drill bits (+/- tolerance) in both metric and imperial
- given a shaft size, apply the same calculation above, but take into account variables like: loose fit, shrink fit, interference, etc...
- same calculation as above, but start with screw of particular type
- calculate the best matching reamer as a second opp
- calculate target hole size assuming third opp ID grinding
- calculate speeds and feeds for all above
- given target material, provide hints about recommended tooling, possibly even pull up catalog P/N for various manufacturers

This is just an example, to give idea about the concept....would cover many different types of simple calculations that are normally manual lookups in a book, or calculator with multiple steps, etc...

The tool would have the full backing of lots of data (a la machinery handbook) so you no longer have to look up in a book. The tool would be highly graphic and interactive, and would allow for easy dragging around, and slide bars for selecting sizes, tolerances, etc....

My question/What I'm looking for is:
- does this sound interesting and useful?
- what calculations would you like to see?

For sure I would play with it as a lot of decision making and processes design hinges on tolerance stack up - between processes. And often working backwards from highest / tightest / most difficult tolerances and features.

Folks that work in shops that have well established processes probably don't need to fiddle with such things. Same with very experienced more old school transitioned to new school machinists.

IMO / IME ~ just do it, (Don't ask "Permission" to 'Do" "It".

And then see how folks might relate to it or use it.

If it's a good idea it's because YOU made it a good idea ~ but I think / almost know, that you already know that

"Execution" is more than half the battle (as you also know).

and again depends how you want to spend / use your time.

Selfishly (in a good way) - what would you get out of it ?


________________________________________________________________________


Something real world on Wire EDM tolerance stack up + complexity of different materials + actual geometries + (machines) might be useful / interesting challenge.

If you don't know all that stuff off the top of your head or a simple chart...you need to go back to machinist school.

Alternate response:
If you don't know all that stuff off the top of your head or a simple chart...you may not be a machinist. (Think "you might be a redneck)

drcoelho said:

I've been considering building a web based CAD like tool that would make various calculations easier for day to day machinists and engineers.

Click to expand...

Look up Tim Markoski, "Machinist Toolbox". Tim was a talented five-axis programer and post-processor writer. Should give you some idears.

Personally, I don't like online stuff, prefer to have something local.

plastikdreams said:

If you don't know all that stuff off the top of your head or a simple chart...you need to go back to machinist school.

Alternate response:
If you don't know all that stuff off the top of your head or a simple chart...you may not be a machinist. (Think "you might be a redneck)

Click to expand...

Its not a question of knowing the stuff, its more a matter of convenience and speed. If what had a little app that saved a minute here and a minute there, seems like that could add up. I'm sure the most experienced machinist still has to go to a book once in a while or do a lookup online. This just compresses the time to do so.

What I'm hoping is that folks that are very experienced as machinists and suggest little aids that would save them a little time here and there....

EmanuelGoldstein said:

Look up Tim Markoski, "Machinist Toolbox". Tim was a talented five-axis programer and post-processor writer. Should give you some idears.

Personally, I don't like online stuff, prefer to have something local.

Click to expand...

Ya, was thinking it might be a hybrid cloud/smart phone hybrid so one could do local stuff as well...

EmanuelGoldstein said:

Look up Tim Markoski, "Machinist Toolbox". Tim was a talented five-axis programer and post-processor writer. Should give you some idears.

Personally, I don't like online stuff, prefer to have something local.

Click to expand...

Good lead on Tim Markoski, thx. I'm thinking of something a little more interactive and with some limited CAD features....sorta something that sits between a calculator time tool (like Machnist Toolbox) and a full blow CAD application. Allows for quick and dirty what-if analysis, taking into account tolerances, and handles multi-step processes, and has a lot of data behind the scenes to just make things easy.....

cameraman said:

For sure I would play with it as a lot of decision making and processes design hinges on tolerance stack up - between processes. And often working backwards from highest / tightest / most difficult tolerances and features.

Folks that work in shops that have well established processes probably don't need to fiddle with such things. Same with very experienced more old school transitioned to new school machinists.

IMO / IME ~ just do it, (Don't ask "Permission" to 'Do" "It".

And then see how folks might relate to it or use it.

If it's a good idea it's because YOU made it a good idea ~ but I think / almost know, that you already know that

"Execution" is more than half the battle (as you also know).

and again depends how you want to spend / use your time.

Selfishly (in a good way) - what would you get out of it ?


________________________________________________________________________


Something real world on Wire EDM tolerance stack up + complexity of different materials + actual geometries + (machines) might be useful / interesting challenge.

Click to expand...

Definately thinking along the lines of tolerance stack-up. Good suggestion here....

drcoelho said:

Definately thinking along the lines of tolerance stack-up. Good suggestion here....

Click to expand...

Cad systems doo tolerance stack up now.

Your not working with a CAD system, just a calculator.

Maybe get with DMFTom and get a spreadsheet going.

digger doug said:

Cad systems doo tolerance stack up now.

Your not working with a CAD system, just a calculator.

Maybe get with DMFTom and get a spreadsheet going.

Click to expand...

I guess the question is this: is there a need for an intermediate tool somewhere between a calculator and a CAD tool that does more than a calculator but is simpler/trivial to use compared to CAD tool? The problem with calculator is that you have to use it repeatedly as you progress through you steps, and it is hard to do "what if" analysis. For example, imagine one were doing tolerance stacking and wanted to work backwards, e.g. target tolerance, then determine what tolerance was required for previous steps. The tool I envision would allow for multiple back to back calculations, one for each op, but without requiring one to build a full CAD model of the part.

I do that sort of thing with a spreadsheet, one that grows more tabs every year! I don't like on-line tools that may or may not be there next week and aren't usually quite right for what I want to do.

digger doug said:

Cad systems doo tolerance stack up now.

Your not working with a CAD system, just a calculator.

Maybe get with DMFTom and get a spreadsheet going.

Click to expand...

Get Thermite involved. He can put it into language that ANYONE can understand.

drcoelho said:

I guess the question is this: is there a need for an intermediate tool somewhere between a calculator and a CAD tool that does more than a calculator but is simpler/trivial to use compared to CAD tool? The problem with calculator is that you have to use it repeatedly as you progress through you steps, and it is hard to do "what if" analysis. For example, imagine one were doing tolerance stacking and wanted to work backwards, e.g. target tolerance, then determine what tolerance was required for previous steps. The tool I envision would allow for multiple back to back calculations, one for each op, but without requiring one to build a full CAD model of the part.

Click to expand...

Just a small tangent, (I don't really have time to burn today ),

So the Studer sales guy I talked to really tried to upsell me on machines when I was really looking for pretty much their nice hands on base model.

Said sales person being well intentioned was trying to tell me that with the diameter of parts I was talking about that the machine I wanted was the WRONG one and I needed something with much more torque and grunt etc.

- Even though he didn't ask what my auxiliary processes were and what was left to "finish" at or how much material had to be removed or what hardnesses etc.

He was basing things on something that must have been more related to plunge grinding and or "Peel" grinding etc. (+ reasonable arguments for process repeatability and cited future and current 'Skills gap" ).

So using a cylindrical grinder the "Wrong way" is inefficient wasteful and destructive (for sure),

But OTOH with some (more conventional) grinding equipment (in house) you can in some ways get away with 'Lesser" CNC "machining" - machines (depending on materials ) you are working with. [Lesser in the sense of less accurate or super rigid. ]. So if a part (initially) coming off a "Milling machine" only has to be within 4 thou and critical surfaces and bores are "rendered" to much tighter tolerances using other grinding processes AND metal finishers and other treatments are going to be part of the process anyway , then the high $ ultra capable machines can or could be utilized for something else or not purchased / or even financed / leased (this round). [i.e. metal finishing - tumblers etc. - means the expensive tooling and synthetic granite -made of machine - is not necessary for insanely good surface finishes straight off a machine ? . ].

The problem is in a lot of cases is spindle runout and quality of tool holders for every machine in a chain of processes + type and architecture of machines and spindle design (hydrostatic + "V" ways or one V and one F with rolling element bearings on a grinding spindle ? ) etc.

Someone would leave 10- 15 thousands whereas some-one else might leave 4 to 2 thousands for the next process ?

It's almost like time and actual spindle capability (spindle curves, HP, rpm and torque ) or even wheel used and method of grinding or machine capability seems it might have a disproportionate effect on how one really goes about "It" - whatever one is doing.


It's interesting when I read Frank Mari's comments over time how he and Maritool go about things in terms of process methodology and approach. Techmaster [Can't remember which good Taiwanese brand he first used ] vs $2M Studer and other DOOSAN mill turn machines he / they have now. i.e. nominally and intrinsically the same process steps but executed quite differently.

Just spit balling ~ Time vs best / optimized process design vs. what's actually doable with the machines you have and how + skill level and experience. materials and condition / heat treat being another level of complexity.

___________________________________________


Software a bit like an "Expert system" and modifiable templates with the capability to have changes "ripple through" in a bi-directional way could be handy. And flag or trigger certain threshold criteria. Or try to map or model how things change if you impose tighter tolerances - does the whole process (chain) fall apart ? + ultimate costs - short term and long term ?

It's a little different for Job shop vs. in-house manufacturing.

Job shops already know what they can do, and in most cases ( hopefully) - how.

Small manufacturers focusing on very specific new product and design / engineering needs - may not be so "Obvious" - for many reasons. - including not knowing "How" - yet.

Easy when you know how when you have a processes chain that's super dialed-in and optimally tweaked until someone suddenly one day declares that it's totally obsolete lol.

cameraman said:

Just a small tangent, (I don't really have time to burn today ),

So the Studer sales guy I talked to really tried to upsell me on machines when I was really looking for pretty much their nice hands on base model.

Said sales person being well intentioned was trying to tell me that with the diameter of parts I was talking about that the machine I wanted was the WRONG one and I needed something with much more torque and grunt etc.

- Even though he didn't ask what my auxiliary processes were and what was left to "finish" at or how much material had to be removed or what harnesses etc.

He was basing things on something that must have been more related to plunge grinding and or "Peel" grinding etc. (+ reasonable arguments for process repeatability and cited future and current 'Skills gap" ).

So using a cylindrical grinder the "Wrong way" is inefficient wasteful and destructive (for sure),

But OTOH with some grinding equipment you can in some ways get away with 'Lesser" CNC machines (depending on materials ) you are working with. [Lesser in the sense of less accurate or super rigid. ]. So if a part only has to be within 4 thou and critical surfaces and bores are "rendered" using other grinding processes AND metal finishers and other treatments are going to be part of the processes the high $ ultra capable machines can or could be used for something else or not purchased / financed this round. [i.e. metal finishing - tumblers etc. - means the expensive tooling and synthetic granite -made of machine - is not necessary.

The problem is in a lot of cases is spindle runout and quality of tool holders for every machine in a chain of processes.

Someone would leave 10 thousands whereas some one might leave 2 thousands for the next process.

It's almost like time and actual spindle capability (spindle curves torque ) or wheel used and method in grinding or machine capability seemed it might have a disproportionate effect on how one really goes about "It" - whatever one is doing.


It's interesting when I read Frank Mari's comments over time how he and Maritool go about things in terms of process methodology and approach. Techmaster [Can't remember which good Taiwanese brand he first used ] vs $2M Studer and other DOOSAN mill turn machines he / they have now.

Just spit balling ~ Time vs best / optimized process design vs. what's actually doable with the machines you have and how + skill level and experience. materials and condition / heat treat being another level of complexity.

___________________________________________


Software a bit like an "Expert system" and modifiable templates with the capability to have changes "ripple through" in a bi-directional way could be handy. And flag or trigger certain threshold criteria. Or try to map or model how things change if you impose tighter tolerances - does the whole process (chain) fall apart ?

Click to expand...

Yes, you are getting to where I was thinking of going....step 1 <--> step 2 <--> step 3 <--> etc.... and all the steps have data backed dynamic calculations that can save you time having to go lookup in a book somewhere.....much like a virtual sandbox for machining a part, you get to try out operations without doing them and understand the impact of some step in the middle of your process. AND lots of templates for just about any process someone could possibly need.

drcoelho said:

Yes, you are getting to where I was thinking of going....step 1 <--> step 2 <--> step 3 <--> etc.... and all the steps have data backed dynamic calculations that can save you time having to go lookup in a book somewhere.....much like a virtual sandbox for machining a part, you get to try out operations without doing them and understand the impact of some step in the middle of your process. AND lots of templates for just about any process someone could possibly need.

Click to expand...

Pardon me for mentioning this, but ^ is certainly not

I've been considering building a web based CAD like tool that would make various calculations easier for day to day machinists and engineers. This would be something short of a full CAD application, but would be a time saver for simple calculations.

Click to expand...

Plan B ^^ sounds more like a desktop computer application ....

I think it would be far more useful to develop a direct plugin for cad/cam programs, in my mind it sounds like a quick "canned cycle/simple cycle" builder.

I did something similar to this for simple operations like squaring blocks, drilling, tapping, reaming and counterboring. It definitely saves time on the simple stuff.

For example, if I pick on a block (lets say its called 1-4-20-tap) it will check for that name in a library of cycles and so long as it finds the cycle it will program everything in that cycles info automatically. (in the case of a 1/4-20 tap thats a spot, drill, spot chamfer, and rigid tap)

TheBigLebowski said:

I think it would be far more useful to develop a direct plugin for cad/cam programs, in my mind it sounds like a quick "canned cycle/simple cycle" builder.

I did something similar to this for simple operations like squaring blocks, drilling, tapping, reaming and counterboring. It definitely saves time on the simple stuff.

For example, if I pick on a block (lets say its called 1-4-20-tap) it will check for that name in a library of cycles and so long as it finds the cycle it will program everything in that cycles info automatically. (in the case of a 1/4-20 tap thats a spot, drill, spot chamfer, and rigid tap)

Click to expand...

Thats an interesting idea, I was thinking of a tie-in to CNC.....