4 answers
2
Questions
Asked
618 views
As a mechanical engineer what do you have to do to insure you don't make any mistakes in your final model?
#engineering #mechanical-engineering
Login to comment
4 answers
Updated
Xinrui Alice’s Answer
Hi Nerissa,
The typical process I've seen is as the following:
-Problem statement. What is the purpose of what you're designing? This is important to define early to avoid this https://images.app.goo.gl/LVbJkY9tDtg5u5gb7. Keep in mind who will use your design and make sure to consult with them to make sure your problem statement is accurate.
- Prototype your design. Put your idea on paper or computer. You'll often find issues that you did not anticipate in your mind, which is totally okay. This is the best learning opportunity to work with others and learn to solve problems. If you have access to rapid prototyping such as 3D printing, this is where you can try out your design to see if it does what you want it to do. However, 3D printed results do not 100% represent what you will get with a different manufacturing process or material. Review end results with other engineers.
- Perform simulations. What are the stresses or motions your design might see? Simulate where stress concentrations are when it bends, or how much force it takes to break. This is great if you don't have access to rapid prototyping, but keep in mind the simulation is only as good as the information you provide. Review end results with other engineers.
The takeaway message I want to emphasize is to get feedback from people you trust as early and as frequently as you can. Don't wait until the very end.
The typical process I've seen is as the following:
-Problem statement. What is the purpose of what you're designing? This is important to define early to avoid this https://images.app.goo.gl/LVbJkY9tDtg5u5gb7. Keep in mind who will use your design and make sure to consult with them to make sure your problem statement is accurate.
- Prototype your design. Put your idea on paper or computer. You'll often find issues that you did not anticipate in your mind, which is totally okay. This is the best learning opportunity to work with others and learn to solve problems. If you have access to rapid prototyping such as 3D printing, this is where you can try out your design to see if it does what you want it to do. However, 3D printed results do not 100% represent what you will get with a different manufacturing process or material. Review end results with other engineers.
- Perform simulations. What are the stresses or motions your design might see? Simulate where stress concentrations are when it bends, or how much force it takes to break. This is great if you don't have access to rapid prototyping, but keep in mind the simulation is only as good as the information you provide. Review end results with other engineers.
The takeaway message I want to emphasize is to get feedback from people you trust as early and as frequently as you can. Don't wait until the very end.
Updated
Alana’s Answer
When you design a model as a mechanical engineer, it is helpful to follow a design for manufacturing guideline to prevent errors that would cause problems in manufacturing. A peer review by other engineers is also helpful. Depending on what is being designed, thermal analysis, stress analysis, and mode effects analysis may also be beneficial.
Updated
Rebecca’s Answer
Can you give more detail surrounding the problem you are solving? What kind of model?
Updated
Dennis’s Answer
Hi Nerissa,
The first thing to note is that we can never be perfect. That is why large companies have service and warranty departments. The more complicated the product is, the more likely it is to have failures. We can never anticipate all the ways a product will be used, and we can't always predict the environmental conditions that will prevail. A good example would be space exploration. We don't have the same conditions here as there are in space, or the Moon, or Mars. Yet, we have been able to capture data and pictures, etc. so that we have a pretty good set of design criteria for some of the environmental conditions. So, now we are pretty confident we can send a rover to Mars to collect pictures and soil samples.
Having a good simulation or model is always a good starting point. Then, subject your design/solution to the extremes that your model or simulation will allow you to go.
If you aren't sure your design or solution process is good, try solving a simpler problem first. I had a professor who explained: "A simple problem is one you know you can get the right answer." We had a complicated multi-degree of freedom problem; we applied the method/process to a simpler one- or two-degree of freedom system to see if the correct solution appeared. Then we could see how it worked for more complicated systems.
Of course, peer review of you work is always good. If you work for a large company, you will have departmental reviews, and possibly a project review of your work product at various times during the product development cycle..
If you are working alone, or as a consultant, you will have to adopt or develop your own set of standards to adhere to. In some cases, there are already industry standards that might apply. For example, SAE has standards for automotive systems - in areas of testing, electronic communication between vehicles and other devices, and the like. ASTM provides specifications on testing materials.
Ultimately, you must test your product in the same way that the potential customer will use it. Simulation gets you to a certain point. But you can't always simulate the heat, vibration, humidity, chemical......(whatever) conditions. Once again - in large companies, durability testing, e.g. testing to failure is part of the design/development process. This is a great way to test your design. Failure is your friend. If something breaks under the controlled conditions of endurance testing, it would very likely break when the customer used it. So, you get a chance to analyze: how did it break; how long did it run before breaking? Was it operating at the expected "worst case" conditions or an overload condition? Can I use a different material or change the shape...? Does the mode of failure alter any of my original assumptions or given design criteria?
Another group process is Failure Modes and Effects Anaylsis (FMEA). You and your peers subject the design to a "what if" process. You study each element in the design and consider ways in which it can fail. And how it might impair the overall system. You rank the severity of the potential failure. Then, go back and figure out if you can reduce the probability of that failure. You know you have a good design when your FMEA score is lower that for the design it is replacing. It is not an easy process, but it is very useful.
of
of
The first thing to note is that we can never be perfect. That is why large companies have service and warranty departments. The more complicated the product is, the more likely it is to have failures. We can never anticipate all the ways a product will be used, and we can't always predict the environmental conditions that will prevail. A good example would be space exploration. We don't have the same conditions here as there are in space, or the Moon, or Mars. Yet, we have been able to capture data and pictures, etc. so that we have a pretty good set of design criteria for some of the environmental conditions. So, now we are pretty confident we can send a rover to Mars to collect pictures and soil samples.
Having a good simulation or model is always a good starting point. Then, subject your design/solution to the extremes that your model or simulation will allow you to go.
If you aren't sure your design or solution process is good, try solving a simpler problem first. I had a professor who explained: "A simple problem is one you know you can get the right answer." We had a complicated multi-degree of freedom problem; we applied the method/process to a simpler one- or two-degree of freedom system to see if the correct solution appeared. Then we could see how it worked for more complicated systems.
Of course, peer review of you work is always good. If you work for a large company, you will have departmental reviews, and possibly a project review of your work product at various times during the product development cycle..
If you are working alone, or as a consultant, you will have to adopt or develop your own set of standards to adhere to. In some cases, there are already industry standards that might apply. For example, SAE has standards for automotive systems - in areas of testing, electronic communication between vehicles and other devices, and the like. ASTM provides specifications on testing materials.
Ultimately, you must test your product in the same way that the potential customer will use it. Simulation gets you to a certain point. But you can't always simulate the heat, vibration, humidity, chemical......(whatever) conditions. Once again - in large companies, durability testing, e.g. testing to failure is part of the design/development process. This is a great way to test your design. Failure is your friend. If something breaks under the controlled conditions of endurance testing, it would very likely break when the customer used it. So, you get a chance to analyze: how did it break; how long did it run before breaking? Was it operating at the expected "worst case" conditions or an overload condition? Can I use a different material or change the shape...? Does the mode of failure alter any of my original assumptions or given design criteria?
Another group process is Failure Modes and Effects Anaylsis (FMEA). You and your peers subject the design to a "what if" process. You study each element in the design and consider ways in which it can fail. And how it might impair the overall system. You rank the severity of the potential failure. Then, go back and figure out if you can reduce the probability of that failure. You know you have a good design when your FMEA score is lower that for the design it is replacing. It is not an easy process, but it is very useful.
Dennis recommends the following next steps: