From DDL Wiki

Client Comments on Report 1

We received your report, and we have some comments and questions. It seems that the existing product is particularly a disaster from a usability and failure perspective and that these are the most obvious areas for improvement. We also are glad to see potential in DFMA and energy use as well. Please respond point by point to the items below. We are looking forward to seeing your ideas in Report 2.

• Executive Summary: Thank you for the clear summary. Please watch typos.
  • We apologize for the errors. This section has been corrected.
• Stakeholder Needs: Thank you for the list, but please comment on the most interesting or critical findings. We noticed that you didn’t list noise or vibration as a need.
  • The critical findings consist of many of the crucial stakeholder needs not being met. These include ease of cleaning/maintenance, effectiveness, safety, reliability, and several others. A more detailed discussion of critical findings has been added to this section of the report. Noise and vibration have also been added to the list of stakeholder needs.
• Use: Excellent detail of use, and very useful observations of issues ranging from difficulty monitoring oil level to difficulty cleaning. Could you compile the major issues into a bullet list? Please provide some type of flow chart to help interpret all of the text. Some questions:
  • A list of the parts of user study has been added to the beginning of this section to help with the interpretation of the text, and the major issues have been summarized at the end of the section.
  • Did you try only 1 of the recipes – might the device work better with the others?
    • We tried several consistencies of batter, and neither worked well, but we did not try all of the provided recipes. However, we looked at customer reviews of the product, and there were people who tried every recipe and got the same poor results as we did during our study, so we did not find it necessary.
  • What is the cause of variance in degree of undercooked output?
    • There were several causes. First of all, the doughnut would occasionally not be flipped properly by the flipper, and would not cook on one side. Also, the doughnut sometimes fell off of the carriage and would be picked up by the next carriage and would then come out overcooked. Additionally, the fluctuation of the operating temperature would cause some variation in how cooked the dough was.
  • The difficulty filling the dispenser is not fully clear from the text.
    • The top of the dough dispenser was strangely shaped, as the dispenser plunger was in the middle of the dispenser cup. Therefore there was only a small amount of clearance between the plunger and the outside of the cup when pouring from a bowl, so it was difficult not to pour dough onto one or the other.
  • You mention that the machine produces one doughnut per minute but later say 20 per hour. Which is it?
    • It is one per minute according to the instruction booklet for the product. This discrepancy has been corrected.
• Assembly: Thank you for the clear pictures.
• Mechanical Function:
  • Please explain the circuit diagram. Some of the symbols are nonstandard and not clear. For example, why are there multiple lines from Conveyor to V-, what trips the switch on the top right of the diagram, what are the empty box symbols, etc?
    • There are multiple lines from the conveyor motor to V- as it is a bipolar stepper motor and most likely utilizing two-wire control.
    • The switch on the top right of the diagram is a temperature switch that turns on or off depending on the temperature of the oil bath.
  • Should the target oil temperature be adjustable by the user? It sounds like the temperature controller is a bang-bang controller – please comment.
    • The target oil temperature is not adjustable by the user, it takes the form of a temperature controller that has only one critical temperature above which it switches off. It is like a bang-bang controller but more similar to a thermostat, where the switch turns the heat on to full until the target temperature is reached, then it turns the heat off until the oil bath falls below the target temperature.
  • Please clarify how switch 2 maintains the circuit until switch 1 is pressed. Are they both on the same component?
    • As it can be seen from the circuit diagram, switch 1 and switch 2 are connected by a wire. They are separate but identical switches. This wire essentially ensures that when switch 1 is turned on, switch 2 is turned off and vice-versa.
• Mechanical Analysis:
  • At what temperature will the outside of the donut "crisp"?
    • This turns out to be irrelevant to our problem, as the "crisping" is a function of both temperature and duration. Given the temperature limitations and the relatively short duration of time, we do not actually expect to see this effect occurring.
  • At what temperature will the oil smoke and/or burn?
    • Canola oil will smoke at 400 deg F. Peanut oil will smoke at 440 deg F. Soybean Oil will smoke at 495 deg F. The analysis has been updated according to this new maximum threshold.
  • Did you leave some safety buffer between the maximum oil temperature you allowed and these problematic temperatures?
    • Yes. We would set the controller to go to 375 deg F when the lowest possible smoke point of the oil is 400 deg F.
  • Which will happen first at high temperatures: crisped donuts, burned oil, or damage to the plastic parts?
    • The temperature at which the oil will begin to smoke is lower than the temperature at which the dough will burn or the plastic will deform.
  • You list some suggested cooking temperatures and times, but how did you use that information?
    • This gave us a benchmark for beginning the analysis. It serves as a reference to validate the thermal FEA results. It was not directly correlated to the results.
  • What temperature does the oil reach currently, and how much more energy will the heating element require to maintain the oil at the higher temperature you suggest, rather than at the current temperature? Assuming the same heat input in both cases cannot be correct.
    • Our data does not currently include the oil temperature, but we can use our knowledge that the heater only operates for about half of the time in its current configuration. To increase the temperature and the cooking rate, we can approximate the heater to be on 100% of the time. These are simplifying assumptions and would require more comprehensive experiments to fully validate.
  • How much time does the dough need to spend at 160 deg F in order to be fully cooked?
    • Only long enough for the chemical change, which for this small mass is negligible. The center of the doughnut will continue to cook even after being removed from the oil, ensuring that once it reaches the desired temperature at the core it will continue to cook through. We do not need to account for this sustained duration at the center.
  • How does changing the time step change your SolidWorks Thermal FEA results?
    • It does not affect the results. It can only increase the resolution of the answer. For instance, shrinking the time step to 0.5 seconds would allow us to know if the preliminary results would yield 41.5 sec/doughnut, 42.0 sec/doughnut, or 42.5 sec/doughnut. This does not increase the accuracy of the results but increases the precision, which given the abundance of inherent assumptions, is unnecessary to the redesign.
  • How long does your redesigned unit take to finish the batch of 60 donuts?
    • 51 minutes.
  • Does the redesign that you suggest pose any additional safety risks or cause any changes in failure modes compared to the current design?
    • No. This does not decrease the structural integrity of the plastic. Should any oil escape the machine, it is possible that it would cause injury, but it would not be significantly different from the risk potential of the original design. We are not adding additional safety concerns. The increased conveyor speed is negligible.
  • Please define all of the symbols used in your notation and be consistent, especially with subscripts.
• Bill of Materials: Thank you for the thorough list. Why do you list several parts under the same part number (e.g. 028 and 036)? Also, please indicate any assemblies with the word “assembly” to avoid confusion (e.g.: conveyor belt).
  • Each component after the doughnut maker was completely dis-assembled in the dissection process was given a distinct part number. There were many sub-assemblies that we were not able to take apart because we would not have been able to re-assemble them after the dissection process. All sub-assemblies were labeled with one part number. However, these sub-assemblies contained multiple parts which each have different functions, are made from different materials and are made using different manufacturing processes, so they are split into multiple lines for one sub-assembly.
• DFMA: Good comments on choice of manufacturing process, but do you also have any comments about DFMA guidelines? What does “a more idiot-proof design” mean – can you be more specific?
  • Comments on all DFMA guidelines are now listed on the DFMA sub-section of Report I. "A more idiot-proof design" has also been clarified to imply that the design should eliminate any assembly mistakes that are caused by ambiguous design. An example of this is there are many different kinds of screws but it is quite easy to try to use the wrong screw. These kinds of error can be eliminated by using more standard screws.
• FMEA: Good identification of major failures (there seem to be many). Should the dough dispenser motor speed be adjustable by the user? We’re surprised it wasn’t designed to be adjustable.
  • As you can see from the revised report, the dough dispenser motor speed is not adjustable and has been included as a failure mode. However, the dispenser motor and the conveyor motor are linked together so changing one the speed of one will change the speed of the other. This would have to be taken into account for a redesign of the dispenser system.
• DFE: The distinction between material flow as direct and gaseous emissions as indirect is not clear - indirect typically refers to up-stream supply chain emissions, while direct refers to the processes being examined directly (such as assembly or use of the product). You mention that metal parts are “designed to be removed” – please comment. It is not clear in your analysis where you calculated use-phase emissions or estimated CO2 tax implications. It appears that you only identified carbon emissions from electricity production used in manufacturing. Use "energy intensive" rather than "energy intense".
  • I should rephrase 'direct' and 'indirect' as tangible or intangible. Indirect wastes, in this case, are defined as those without further consideration by the polluter, whereas direct wastes are those requiring disposal of some sort. The metal is not designed to be removed. This was a unfortunate typo and has been updated in the revised report. My apologies for the error. Use phase emissions are overwhelmingly dominated by the electricity consumption and we have therefore made the approximation that all emissions come from the electricity sector as shown in the EIO-LCA.
• If any of your images, figures, or text were taken from another source, please be certain to provide proper attribution.
  • All images are original images taken by the Doughnut Maker group unless otherwise specified.