Some design questions from a newbie

[andysuth]

Looking at your original design, that would be an easy part to make on a screw machine or lathe (it doesn't have to be CNC). Thickening up the walls is probably necessary, especially to accomodate a real thread and there might need to adjust for the thread cutting operation in the design. Even for production a screw machine would be a good way to get an inexpensive part without a lot of capital investment. If you could consolidate the red and blue parts together (run that blue tapered section up to the inside of the red dome wall) that would further save you money.

It's hard for us to know what you mean about buying an off the shelf part, if you could post a sketch it would be helpful.

The parts shown in image isunmanufacturable, has interferences, suffers massive stress concentrations at root of thread and needs general optimisation.

Coupled with the fact you've probably not got any flatwork (2D drawings) done to tell the blokey making it which dimensions are critical to Quality, and which are just nice to have.

In short, you need to get a proper engineer who understands your purpose to review and design this, otherwise just like this 3D rendering guy (probably not an engineer) the manufacturing guy will do what is easiest for him, not what is best for you.

-Andy Southern.

 

[K.I.S.S.]

Hi COSenna,

In general, I agree with the last post (andysuth), although I wouldn't say your design is unmachinable as a prototype - just extremely tricky, and therefore expensive. It CERTAINLY isn't easy and it CERTAINLY would have to be done on a multiaxis CNC lathe/machining centre unless you want to invest in custom ground tooling. It's possible to toggle the X and Y axis fairly accurately on a manual lathe, but nobody can manually reproduce your specific geometry accurately on a manual machine.
If you do want to go the machining route for a prototype, then add additional clamping material to each component that can be parted off as a last step. But for production, forget it.
As Lochnagar posted, try to be a little more explicit as to the nature of your design - obviously you don't want to let the cat out of the bag, but the more info. the better.
It is possible to use 3D printing for a prototype, but a couple of points - the process you'll require is called DMLS (Direct Metal Laser Sintering). Various alloys are available, but for your stated requirements have a look at AISi10Mg - data sheet is given here http://www.3trpd.co.uk/wp-content/uploads/2014/11/Aluminium-AlSi10Mg.pdf
But this can also be a minefield for you - a fine pitch thread such as you require is beyond the actual (not the stated) accuracy of the process which is 0.1mm at best, so you'll need to consider this real world tolerance when the product is re-modeled.
Also, you'll need multiple support walls that will be required to be grown in conjunction with your models in order to support your components' geometry. Cleaning these up can be a task so time consuming that you'll wish you'd done something else...

Also consider your requirement for anodizing - is it for the look of the thing, or for environmental protection? Some anodizing processes are not U.V. stable, so if they're exposed to the sun, they'll look terrible within a few months.
And most anodizing leaves behind a residue that will clog up a fine pitch thread, requiring cleaning. And lastly on this point, if you require the product to be anodized for environmental protection, anodizing won't provide that for the majority of your internal surfaces, owing to the Faraday cage effect.

And also in relation to what andysuth said, full depth thread roots in thin walled components should be avoided - attempt to have some 'lead' material on the male component.

I'm not trying to put you off, promise! I'm just trying to list some of the complexities involved in what would appear (at face value) to be a simple design.
Think about general things that you might need some help with - what info do we need in order to help?
Are the following important:

Load bearing capabilities
Environmental robustness
Ingress Protection (I.P. rating)
Acoustics
Temperature range
Required quantity
Product longevity
Accuracy
Cost... yeah, right...

Any manufacturing Company in the world will tell you the following;

We do 3 types of job:

Good, fast and cheap - you can choose from any 2 of the above...

If it is good and fast, it will not be cheap
If it is good and cheap, it will not be fast
If it is fast and cheap, it will not be good

K.I.S.S.

 

[andysuth]

+1 on KISS,

He's right, you probably could make it, certainly at prototype stage, but I'd suggest its pointless making it with that thread about about to break through the outer wall.

If you look at the lower right area of the red part in cross section you'll see the thread doesn't even break through to the end face. (hint, the length of red material is larger than the pitch of the thread!)

You should also consider a chamfered lead in for all threads and stress relief at the shoulder end of all threads.

Please don't take these to a manufacturer or prototype house.

Engineers exist for a reason. To go direct from a ID/ stylist to a manufacturer is like diagnosing yourself sick and prescribing yourself medicine without seeing a Doctor.

I've PM'd you and will talk you through things you need to consider, FOC.

All the best,

-Andy.

Hi COSenna,

In general, I agree with the last post (andysuth), although I wouldn't say your design is unmachinable as a prototype - just extremely tricky, and therefore expensive. It CERTAINLY isn't easy and it CERTAINLY would have to be done on a multiaxis CNC lathe/machining centre unless you want to invest in custom ground tooling. It's possible to toggle the X and Y axis fairly accurately on a manual lathe, but nobody can manually reproduce your specific geometry accurately on a manual machine.
If you do want to go the machining route for a prototype, then add additional clamping material to each component that can be parted off as a last step. But for production, forget it.
As Lochnagar posted, try to be a little more explicit as to the nature of your design - obviously you don't want to let the cat out of the bag, but the more info. the better.
It is possible to use 3D printing for a prototype, but a couple of points - the process you'll require is called DMLS (Direct Metal Laser Sintering). Various alloys are available, but for your stated requirements have a look at AISi10Mg - data sheet is given here http://www.3trpd.co.uk/wp-content/uploads/2014/11/Aluminium-AlSi10Mg.pdf
But this can also be a minefield for you - a fine pitch thread such as you require is beyond the actual (not the stated) accuracy of the process which is 0.1mm at best, so you'll need to consider this real world tolerance when the product is re-modeled.
Also, you'll need multiple support walls that will be required to be grown in conjunction with your models in order to support your components' geometry. Cleaning these up can be a task so time consuming that you'll wish you'd done something else...

Also consider your requirement for anodizing - is it for the look of the thing, or for environmental protection? Some anodizing processes are not U.V. stable, so if they're exposed to the sun, they'll look terrible within a few months.
And most anodizing leaves behind a residue that will clog up a fine pitch thread, requiring cleaning. And lastly on this point, if you require the product to be anodized for environmental protection, anodizing won't provide that for the majority of your internal surfaces, owing to the Faraday cage effect.

And also in relation to what andysuth said, full depth thread roots in thin walled components should be avoided - attempt to have some 'lead' material on the male component.

I'm not trying to put you off, promise! I'm just trying to list some of the complexities involved in what would appear (at face value) to be a simple design.
Think about general things that you might need some help with - what info do we need in order to help?
Are the following important:

Load bearing capabilities
Environmental robustness
Ingress Protection (I.P. rating)
Acoustics
Temperature range
Required quantity
Product longevity
Accuracy
Cost... yeah, right...

Any manufacturing Company in the world will tell you the following;

We do 3 types of job:

Good, fast and cheap - you can choose from any 2 of the above...

If it is good and fast, it will not be cheap
If it is good and cheap, it will not be fast
If it is fast and cheap, it will not be good

K.I.S.S.

 

[Stew F]

First up, apologies for the long winded reply.

I think the person who did the design work on that was only getting it to a concept model state. There is a lot of detailing to be done on it. It's very doable as it is, but you will only get what is represented in the model; essentially, a prototype with little or no functionality. As others have said, maybe consider a 3D printed part. But not off that model.

Agree with others here. If you tell us the basics of what it's supposed to do we can offer some advice on how to go about getting a detailed design nailed down. Things like a) what it does (not in any technical detail) & b) where it will be used (environmental considerations ' anodizing may not be the best surface treatment). Is Aluminium the best material? Can it be made of something else, like some type of rigid plastic?

Design aside, then YOU need to consider the commercialisation of your idea. 1. What is your brand? 2. What is the widget called? 3. Who is your target consumer? 4. What sort of quantity do you expect to sell (as opposed to what you might ACTUALLY sell)? 5. Can you patent this idea? (there is a cost). 6. where will you get it made? What are the manufacturing costs? What are the minimum quantities you need to order in each production run? Be careful going to "cheap" manufacturing countries. Some of them have no hesitation in taking a good idea and making it theirs. Once you have nailed all this you need to work out what your return On investment is, and then decide if it is worth the effort (and expense).

Be aware that getting a detailed design onto proper 2D working drawings is not a cheap exersise in itself. I'd expect to spend ~30 to 40 design hours, plus additional time to to do general things like make up a drawing template with your company details (unless you have one), design reviews, documentation, ground work on material and surface treatment selections, and finally, once you are happy with everything, revising the drawings to IFC (Issued For Construction) and issue the drawing pack. There's probably another 2 days in that. So all up, you're looking at 50 - 60 hours for the designer. I don't know what they charge where you are located, but here that would cost you around $4500 - $5000 plus tax. If you are prepared to invest that kind of money in your idea then I say run with it.

Getting "a mate" to do this for you is OK, but when you've made your millions he'll want something. After all, without his work, you haven't made anything. If you get my drift.

Feel free to PM me if you want any more info/advice.

Stew F
South Australia

 

[K.I.S.S.]

I've been lurking around this site for about a year now, and I have to say that this is by far the most collaborative and constructive effort I've seen on it - is this the start of the first 'cloud' design based on free input?
It would be cool if it was - everyone loves a 'widget'...
+2 on Andy - in my defence, I was actually the first to comment on the inadvisability of the thread as it currently is, and to be fair to the person who modeled it, it looks like it was done using SW which still to this day automatically starts a helix/spiral path at 135 degrees (for reasons unknown and unknowable to myself...), which means that any available section view will not accurately reflect the start/end point of the revolve/cut revolve, so given the projected theoretical pitch, it could actually be, and probably is, a full length thread.

COSenna, as you've stated, you're not a designer or engineer, so you're probably not aware of FEA/FEM analysis - this is a method we use to basically break down any design into the smallest relevant measurable elements and interactions. It's full name is Finite Element Analysis, and we use it to study small things like stress, interference, friction and geometry at a very small level. It's a very powerful tool and the results are frequently surprising - but for it to be effective, it's important that any CAD model is constructed with this in mind. In other words, its about how the CAD model should be built to have maximum flexibility in order to allow a small change to be made without 'throwing out' any major features in it.
I could, but won't, go on for a long time about this - just keep it in mind with regard to a 3D model of your part.

The following comment has absolutely no relation to the people on this site who wish to assist you (myself included), it's just something I saw a few weeks ago - "If you think a professional is expensive, wait until you've tried an amateur..."

K.I.S.S.

 

[Dave Archer]

Seems to me, this is out of one of those college class "Design something that is impossible to make" competitions !

"I have contacted a few CNC machinists in the US and China but have not received any feedback yet."

No surprise there !!!
No one will touch it !

To be honest, as an ex-cnc lathe programmer and operator, I can tell you that those companies can instantly see that it is an appalling amateur design and will fall apart as soon as any threading cuts are attempted, apart from the struggle to actually hold the parts in the lathe.

A special threaded jig would need to be made to make the green part.

Making that parallel hand knurl around the top cap would cost too.

...... And you could not afford to pay for wasting their time anyway !

These parts as shown are certainly injection moulded designs in plastics, you would never find these as machined parts.

If you want nice shiny anodised parts, you need machined bar stock 6062 type alloy or similar.
Anodising does not like cast alloy...even polished cast alloy.

It would be interesting to know what this device is even supposed to do !

I'm a Solidworks designer, ( my job ) send me the drawings and I will sort out the disaster, I will do it for FREE.
( Because I am interested ) !

Dave

 

[COSenna]

Sorry for the late reply, busy as ever lately.

Obviously I can see that giving up what the product's intended use is will help many of you help me. I do not though, as stated by K.I.S.S. want to "let the cat out of the bag." Although some might find my idea ridiculous or stupid, I am quite confident it will make me money. That being said, I cannot state exactly what it is. I can however tell you that this will be used for smoking certain plants, either tobacco or marijuana. This is why I was looking towards anodized aluminum. I read online that it is the best option for aluminum as it relates to smoking. I also picked aluminum instead of steal because I need this to be as light as possible, while also being somewhat rugged. Those are my main concerns. Does anyone have any other materials in mind for such purposes?

I am a graphic designer by trade. So I have all my branding finished. I also will have a website once I figure out if this is possible with my budget. This part of the project only costs me time, fortunately.

Speaking of budget, this is why I went with an amateur in the first place. I am trying to keep this side of the project as low as possible so I can have more on the production side. I see now, with all the replies, that this was obviously a mistake! Skill labor is cheap, and cheap labor isn't skilled, right?

Going forward I am looking for someone who can help me model this (correctly) but won't shatter my piggy bank! I know this may be a shot in the dark, but I always hope for the best.

Here is a new drawing of the parts. The "washer-like" piece will be placed on the bottom of the two threads.

Thanks again fellas!

 

[Dave Archer]

I've been lurking around this site for about a year now, and I have to say that this is by far the most collaborative and constructive effort I've seen on it - is this the start of the first 'cloud' design based on free input?
It would be cool if it was - everyone loves a 'widget'...
+2 on Andy - in my defence, I was actually the first to comment on the inadvisability of the thread as it currently is, and to be fair to the person who modeled it, it looks like it was done using SW which still to this day automatically starts a helix/spiral path at 135 degrees (for reasons unknown and unknowable to myself...), which means that any available section view will not accurately reflect the start/end point of the revolve/cut revolve, so given the projected theoretical pitch, it could actually be, and probably is, a full length thread.

COSenna, as you've stated, you're not a designer or engineer, so you're probably not aware of FEA/FEM analysis - this is a method we use to basically break down any design into the smallest relevant measurable elements and interactions. It's full name is Finite Element Analysis, and we use it to study small things like stress, interference, friction and geometry at a very small level. It's a very powerful tool and the results are frequently surprising - but for it to be effective, it's important that any CAD model is constructed with this in mind. In other words, its about how the CAD model should be built to have maximum flexibility in order to allow a small change to be made without 'throwing out' any major features in it.
I could, but won't, go on for a long time about this - just keep it in mind with regard to a 3D model of your part.

The following comment has absolutely no relation to the people on this site who wish to assist you (myself included), it's just something I saw a few weeks ago - "If you think a professional is expensive, wait until you've tried an amateur..."

K.I.S.S.

The initial angle has to start somewhere when the function is first opened, but you change it to whatever start angle you like
to start off the thread sketch angle / position.

 

[Dave Archer]

Was a bit of a struggle to understand those drawings but how about this design ?

Thread is 0.040" pitch.

 

[K.I.S.S.]

Hi Dave,
I'm fully aware of the ability to change the start angle of a helix/spiral in SW - I was commenting on whether the original 'modeller' was aware of the start angle - which still begs the question - why 135 degrees, as opposed to 0 or 90, in alignment with predefined reference planes? Just another little SW gripe from me, but I don't want to sabotage this thread through going on about it, so let's just say that there are no 'perfect' parametric modelling programs.