Lecture 15: Test Procedure in Engineering Design

Transcription

1 MECH 350 Engineering Design I University of Victoria Dept. of Mechanical Engineering Lecture 15: Test Procedure in Engineering Design 1 Outline: INTRO TO TESTING DESIGN OF EXPERIMENTS DOCUMENTING TESTS 2

2 Testing & Evaluation within the General Design Process Identify Need -Talk with Client -Project Goals -Information Gathering Problem Definition -Problem Statement -Information Gathering -Design Objectives (quantifiable/measurable) Conceptualization -Brainstorming -Drawing/Visualization -Functional Decomp. -Morphologic Chart Preliminary Design & Planning -Prelim. Specifications -Prelim. Analysis -Decision Making -Gantt Charts & CPM Detailed Design -Detailed Analysis -Simulate & Optimize -Detail Specifications -Drawings, GD&T Prototyping -Prototype Fabrication -Concept Verification Testing/Evaluation -Evaluate Performance -Are Objectives Met? -Iterate Process Steps 2-7 as needed Report/Deliver -Oral Presentation -Client Feedback -Formal Design Report 3 Introduction to Testing Testing is an essential part of the engineering design process. Upon completion of the detailed design stage, the design team must evaluate the design to determine if the design objectives have been satisfied. 4

3 Introduction to Testing Test results can take different forms: Qualitative (value statements, observations) Yes/No Great, Good, Acceptable, Poor Quantitative (numeric, measurable) Meters/second, Kilograms, Newtons Decibels, etc... In MECH350, we are interested in the evaluation of the design objectives. Hence, your tests should generally produce quantitative results. Quantification of test results (with proper testing) allows for designers and clients to agree that the design objectives have been met. 5 Formal Testing Methods: The Scientific Method This is the scientific method/process through which we acquire knowledge and build understanding on how nature works. Hypothesis, Conjecture or Idea Design of Experiment - Experiment Type - Variables - # of Samples Conduct Experiment - Procedure - Data Collection - Other Observations Analysis of Data - Statistical Analysis - Report Results Discussion of, and Reflection on Results - Report Results - New Ideas!!! 6

4 Formal Testing Methods in Engineering Design: Design of Experiments An engineering experiment is used to test the performance of an engineered design, and is similar to the approach used in the scientific method. Define Testing Scope - Goal/Purpose - Expectations - Develop Test Plan Design of Experiment - Experiment Type - Variables - # of Samples Conduct Experiment - Procedure - Data Collection - Other Observations Analysis of Data - Statistical Analysis Discussion of Results - Test Report - Future Considerations 7 Developing a Test Plan A good test plan/procedure must be developed prior to testing. This plan should include the following aspects: Scope: Identify the system and goal/purpose of the test Relevant parameters for the test and justify Expectations of the test (Hypothesis) Administrative Details: Date and Location of Testing Client or Organization that has requested the test Who is conducting the test 8

5 Developing a Test Plan (cont.) Design of the Test (Also know as: Design of Experiment) Type of testing method and its significance List of test apparatus and measurement equip, model #s, etc... Identify dependant and independent variables # of factors considered (single, double (factorial), etc...) Sampling procedure: How are samples obtained? Number of samples Detailed, step-by-step procedure to conduct the test. Conduct/Perform the Test/Experiment Safety precautions Data collection method (written, recorded, digital, etc...) Observation of external factors (temperature, wind, noise, vibration, etc...) 9 Developing a Test Plan (cont.) Analysis of Test/Experiment Data Conclusions Interpretation of data analysis, and other observations Written test report New Hypothesis, leading to re-design of test 10

6 Design of Experiments for Engineers The subject of Design of Experiments is intimately related with statistics and statistical analysis, and is a major area of study. Hundreds of books, and courses are dedicated to the area. Examples where rigorous testing is essential to demonstrate the performance of a consumer product: If you are the engineer in charge, how are you going to demonstrate the product meets performance expectations? How are you going to prove/demonstrate it to the world? Answer: --> Proper Design of Experiment 11 Design of Experiments for Engineers Design of Experiments involves factors such as: Formal Hypothesis Experiment Type: Single- Factor, Factorial (multi-factor) Dependant Variable Independent Variable # of Samples Randomization Blocking Orthogonality Statistical Significance Regression Covariance Analysis 12

7 Design of Experiments for Engineers Formal Hypothesis: A formal statement of what the experimenter expects to find in the data. For example: Treatment of a material sample, will lower the percentage of sample failures under load. Null Hypothesis: A Hypothesis that is stated in a null form (which is more easily testable). For example: There will be no difference in the percentage failures between those samples treated, and those that were not. 13 Design of Experiments for Engineers Independent Variable: Is a controllable experimental variable. By changing/controlling the Independent Variable during an experiment, we seek to see if that contributes to a response in the Dependent Variable. Dependent Variable (Response Variable): Is a characteristic outcome of an experiment on which information is to be gathered or obtained. The outcomes of a true experiment should be random, so a dependent variable is a random variable. There is a probability distribution associated with different kinds of random variables, such as: Normal, Poisson, Binomial, Chisquared, Student t, Beta, Gamma, and many other distributions. 14

8 Design of Experiments Full treatment of the subject of design of experiments is beyond the scope of MECH350, but it is vitally important in the testing phase of engineering design. For further information, consider the following references: [1] C.R. Hicks, K.V. Turner, Fundamental Concepts in the Design of Experiments, Fifth Edition, Oxford University Press, Wikipedia, Design of Experiments, available online: Design_of_experiments 15 Test Reports for MECH350: Good testing documentation serves to show others that your design has met the intended design objectives. The test report should be written in such a way, as to allow an independent, third-party to review the test plan, the experimental design, the collected data, and the data analysis, and allow them to generally come to the same conclusions/test results that you did. A good testing report is similar to a laboratory report, but includes extra elements (next page). 16

9 Test Reports for MECH350: Elements of a Test Report: Scope (see page 9) Administrative Details (page 9) Design of Test (page 10) Procedure (step-by-step of how test is conducted) Data collected Summary if lengthy, with raw data in Appendix. Other noteworthy observations (just state the facts and leave comments/discussion for later). Analysis and Interpretation of Data Summary of Results and Discussion Table format for this is helpful (see next page) Ideas for Further experiments or experimental improvement Conclusion 17 Test Reports for MECH350: Example of: Table Summarizing Results with Discussion Should have four columns: Original objective + it value Actual objective value from testing Declaration of meeting objective Explanation/discussion of result Objective (Target) Test Result Objective Met? Discussion Size (L x W x H) La x Wa x Ha Yes Explain... Weight (10 lbs) 12 lbs No Explain... Noise (73 db) 70 db Yes Explain... Cost ($350) $250 No Explain... Safety (described page x) (described on page y) Yes Explain... Etc... 18