Machine designs tips.

[CORDVI]

Hi, I work making machines and fixture designs. I been thinking on a better way to do my job. We have a lot of problems always we start a new machine and I don't know if that problems are the same for all. That's why i'm asking you for some useful tips to make faster designs and beters. I need to know tips about all the process of maching designs and the design as activity.

I hope that all of your understand my thread, my english is bad.

Thanks a lot.

 

[StefanHamminga]

Could you please describe your typical design process? That would make it easier to spot where you might improve.

Some things that might help:

I design a lot of prototypes (actually, 95% of what I do is design 'new' machines).
The key in my work is communication, and I mean up & down the chain, get the folks that have to build your machine involved early, as they will give you very valuable feedback (and complain a lot less when they actually build it)!

Desiging (for what I do) is all about finding the best compromise between the demands from the client and the demands from the 'work floor', try to find out both and document them.



Stefan Hamminga
Mesken BV
2005 Certified SolidWorks Professional
Mechanical designer/AI student

 

[Barry1961]

As Stefan said communication is key and the earlier the better.

Do not limit your communication to meetings, memos and emails.

1. Having the designer help build a machine will be well worth the time. It is amazing how easy it is to explain how to assemble something when you don’t have to do it. If a picture is worth a thousand words actually going out and doing it is worth a thousand pictures.

2. Learn about the customer’s product or process. While this should be a given it is one of the most overlooked. Often you are having someone design a machine to do something they have absolutely no experience in. The designer will not know enough about the process to know what questions to ask. Getting the customer and future machine operators involved in the design process will give them a sense of ownership.

Often times the plant engineers will give incorrect data for design specifications. This can be because they do not know or they plan to have some current problem solved in near future. Either way if YOUR machine is installed and does not work it is bad news.

3. Learn about your product and process. This is often the most unpleasant part and may be too much for some designers to stomach. What does your company do? What do your salesmen sale?

There is often a huge gulf between design and sales/estimating. As much as they depend on each other a little understanding should be required. This is also true for purchasing.

Barry1961

 

[RobWard]

Consider the future maintenance of any machines.
I always go that extra mile and put enough access panels in so that if I need to go back and change something I can do so without having to destroy the thing.
I hate working on equipment where you have to nearly take an angle grinder to it and cut it all to pieces when all that was neded was for, say, a bolt to be fastened in from a different direction to the one they chose...

"I love deadlines. I love the whooshing noise they make as they go past." Douglas Adams

 

[DesignerMike]

I wholeheartedly agree on the communication issue.

When I started my last job, I was very interested in how they actually built the equipment (mostly since I had been doing a lot of fabrication in the years before I moved more into the engineering department). I spent a couple hours a week on the floor and quickly became friends with some of the welders and assemblers. I guess I kind-of spoke their language, and they were extremely excited that someone from engineering cared about what they thought.
Needless to say after about 3 months the number of ECNs generated from my jobs was down to almost 0, and there were other designers with 15 years of experience at this plant having 10 plus ECNs per job. Management actually noticed, and started to push all the designers and engineers toward spending a little more time on the floor. I'm sure it was a combination of understanding the fabrication process better, AND whenever the guys on the floor would come across a minor drawing error (like I only had 3 nuts called out on the BOM instead of 6) they would just "handle it" instead of requiring paperwork, new drawings, ECN, and alert purchasing etc.

The customer communication is more difficult. My current job gives me a lot more interaction with the customers, and I spend much more time in their plants. I spend quite a bit of time modifying new equipment before it even runs. It is usually just poorly designed because either the manufacturer didn't understand the process (or didn't care to). It's pretty bad when you look at a new conveyor system, and see perfectly good product on the floor in piles. VERY educational to figure out how it got there. Customers don't really care how it got there, they just don't want to see more of their money going down the drain.

 

[Barry1961]

Another thing that I have seen cause a lot of problems is that some people will not accurately draw or document the machine.

Some of this may be for a false sense of job security, “I am the only one who knows how it works”.

Some of this may be holdovers from drafting board days when assembly drawings were cartoons.

Some really believe the accuracy of the drawing does not matter as long as they can make the machine work.

Whatever the reason I would fire these people very quickly. You can not compete today by drawing everything from scratch. You must have accurate drawing libraries of proven components and assemblies. If you have CAD libraries that are full of inaccurate drawings you would be better off with napkins sketches.

I have had to deal with a lot of machines where the drawing and BOM that had very little in common with the actual machine. Of course the person who designed the machine was no longer working there. So now if the customer needs a modification you have to go to the machine and make a drawing accurate drawing and BOM.

Barry1961

 

[MintJulep]

In addition to the excellent advice above:

Machines are complex systems, and need to be designed as a system. The mechanical guys and the electrical and software guys need to talk during the design, not after.

How will the machine be controlled? What inputs will the software require to accomplish that control? What sensors are needed to provide those inputs? Where and how do the sensors get mounted?

Write the system functional description FIRST. Start your FMECA early, and keep it up to date as the design progresses. Use the FMECA as a tool to develop a meaningful test program.

Make use of past experience - both good and bad - but not blindly.

 

[CORDVI]

Thanks for the excellent advice. I really apreciate.

Here I will describe the steps of the process to build a machine in my job.
steps:
#1 Go to the client to review her needs, and after that revise some concepts to start the design.
#2 Start the machine or fixture design.
#3 After that, we must buy some mechanical parts, as pneumatic actuators, valves, sensors, bushings etc.
#4 Make drawings for part that need to be machining.
#5 Start whit the machinig process of parts, like milling, turning and so on.
#6 Assembly the machining parts.
#7 Make the control program (most of our machines are controlled whit PLC)

All of the steps above are the process fo making machines in my actual job.
The steps 5, 6 and are made it whit the help of technicians, and the step 7 must be conclude by one control engineer.

 

[CORDVI]

Continuing whit more questions.

MintJulep. Right now whe dont uses FMECA. How can I implement this to my process of machine designs.

I think tha we cant make some system designs almost the same for all machines, to save time. This can be done only if the system are almost the same.

Other Question. How to minimize the mistakes.

Thanks a lot.

 

[MintJulep]

Cordvi,

Where do you document how you will meed the client's needs?

Where are your design reviews?

A FMECA is a structured exercise in critical thinking. Look at each component in the design and list all of the things that could go wrong with it and the consequences of each thing going wrong. Then use that information to design out as many of the weaknesses as possible. There are many good books on the subject.