How do Engineers Calculate the Dimensions of their Design Ideas?

[Al-Layth]

TL;DR Summary

How do Engineers Calculate the Dimensions of their Design Ideas?

You have a design problem statement, which gives you numerical specifications for performance, spatial constraints etc, and you have finally come up with some mechanism ideas you synthesised into a machine concept.

You’ve sketched your design’s parts and the final assembly.

So: How do you now go about calculating the dimensions of your undimensioned design?

 

[anorlunda]

It may be easier to consider a specific case. Take a crane for example. How would you go about setting the dimensions? What would the requirements say about dimensions?

 

[jack action]

gives you numerical specifications for performance

Those numerical specifications must translate into dimensions somehow.

To give you a hint about the problem proposed by @anorlunda , everything is centered on the weight hanging at the end of the rope. What are the numerical specifications for what you hope to achieve with this weight?

 

[russ_watters]

You have a design problem statement, which gives you numerical specifications for performance, spatial constraints etc, and you have finally come up with some mechanism ideas you synthesised into a machine concept.

You’ve sketched your design’s parts and the final assembly.

So: How do you now go about calculating the dimensions of your undimensioned design?

I don't really understand the question. The critical dimensions are defined from the device requirements/specifications.

 

[Lnewqban]

Summary: How do Engineers Calculate the Dimensions of their Design Ideas?

You have a design problem statement, which gives you numerical specifications for performance, spatial constraints etc, and you have finally come up with some mechanism ideas you synthesised into a machine concept.

You’ve sketched your design’s parts and the final assembly.

So: How do you now go about calculating the dimensions of your undimensioned design?

Taking as example a handheld pump that will be used to suck diesel from a car's fuel tank.

You have a problem to solve, and the solution should adjust to the constrains of that problem:
The arms muscular effort of one single person, the diameter of the hose entering the fuel tank, the height difference between tank and accumulating container, estimated time for the operation.

Then, you need to analyze the allowable cost of your device, including available materials, prefabricated parts, manufacturing processes, number of units to be produced.

For parts that are not commercially available and that you need to get manufactured, you will need to calculate the resistance and endurance of the parts, based on the less costly material and fabrication methods.

 

[BillOnne]

These days, an engineer would probably use some sort of computer aided design (CAD) program. There are a number of them. They do a variety of calculations on demand.

Such software is typically fairly expensive. The more features and the smaller the industry, the more expensive it is likely to be. Plus, there may be special customization for the specific task, project, or device being designed.

The effort up front includes learning to use the software and creating the design in the software. You also need a fairly good computer to run it and use it. They typically require a butt load of RAM, hard disk space, CPU power, and you will want a large hi-rez monitor. Plus you probably want some largish printers for printing blueprints.

Once you have that, they can usually do a bunch of interesting useful things. For example, they can calculate the size, mass, etc., of the system. Depending on the program, they may give you help with designing the process of building the system. For example, there is the "ship in a bottle" problem. If something needs to go inside something else, you need to be sure there is a way to get it there. Depending on the type of thing being built, your CAD program may give you help with assembly sequence, bill-of-lading telling you what you need to order, and a bunch of other stuf.

They can also do things like creating blueprints for a variety of needs. Such as construction, maintenance, commissioning, deconstructin, and decommissioning.

The up-front cost is not trivial. But being able to do all that stuff from the basic design is hugely valuable. In my industry, I frequently have to look at blueprints that were hand drawn in the 1970s. And there have been several modifications of the system since it was built, with the blueprints not being updated. Or being updated by hand. With low-rez scans that look like somebody raised a family of parrots on it. If it was all in a CAD program, I could look at the new drawings, even look at the design history.

 

[DaveE]

This is the art of a practicing engineer in the real world. The answer depends on way too many things to list and varies between disciplines, products, markets, etc. The job of an engineer is to do a "good enough" job quickly. Yes you do some physical science stuff to constrain the design and define "good enough", but you are likely to have other sloppy requirements, like aesthetics, business constraints, coworker personalities, consumer/customer preferences, previous design experience (like proven reliability), reduced/optimised inventory, supplier/purchasing issues, regulatory and documentation issues, etc. And, in my experience, the elephant in the room; time to market and NRE cost.

Personally, in >30 years of EE product design experience, I have never done a design that I thought couldn't be improved, given more time and money, and seldom had the opportunity to do it completely the way I wanted. You can't work forever on anything. You have to satisfy diverse requirements, some of which you may not agree with, some of which may conflict with each other.

For example, if your design mixes SAE and metric hardware, EVERYONE will hate you, even if your computer says it's better that way.

 

[jrmichler]

This is the art of a practicing engineer in the real world.

It is an art, and the art is based on 4 to 5 years of college engineering courses plus practical experience. The engineer will sketch a concept, calculate the forces on each part, then calculate the stresses and deflections of each part. Those stresses and deflections are compared to allowable stresses and deflections that the engineer previously calculated.

Then the result is compared to the product specification. If it's good enough, the design is detailed and released. If not, it is iterated. Repeat until it is good enough or management sends down an ultimatum to release the design.

Plus everything that @DaveE said in Post #7.

 

[Ranger Mike]

How do Engineers Calculate the Dimensions of their Design Ideas?

Reference:

It is an art, and the art is based on 4 to 5 years of college engineering courses plus practical experience. The engineer will sketch a concept, calculate the forces on each part, then calculate the stresses and deflections of each part. Those stresses and deflections are compared to allowable stresses and deflections that the engineer previously calculated.

Then the result is compared to the product specification. If it's good enough, the design is detailed and released. If not, it is iterated. Repeat until it is good enough or management sends down an ultimatum to release the design.

Plus everything that @DaveE said in Post #7.

real simple
Metrology is the science of Measurement
Measurement is the language of science
learn about measurement so you can ask questions that lead to answers