REPORT OF CFD LAB 1
Number of Students
Total 24
Submitted 23
Not Submitted 1
1 1
0 1
3 6
11
0 2 4 6 8 10 12
65~70 70~75 75~80 80~85 85~90 90~95 95~100
Grade Distribution
91.1 89.8 89.8
95.5 93.9
89.9 91.7
10.0 13.5 8.9
17.3
7.8 9.1 8.0
0 20 40 60 80 100
2009 2010 2012 2013 2014 2015 2016
Comparison with Previous Years
Average SD
23
1
0 5 10 15 20 25
CFD Lab 1 CFD Lab 2 CFD Lab 3 CFD Lab 4
Submission History (2016F)
Submitted Not Submitted
91.7
8.0
0 20 40 60 80 100
CFD Lab 1 CFD Lab 2 CFD Lab 3 CFD Lab 4
Comparison of CFD Labs (2016F)
Average SD
Average: 91.7
Standard Deviation: 8
1. Common Mistakes
a. A couple of students didn’t follow the report format
b. For the last question of problem 9.2, few students compared % error with the uncertainty which is divided by nothing (refer the formula inside the cells in the excel file). To compare the error and uncertainty, the denominator should be matched.
c. For problem 9.3, many students answered that the grid sets containing grid #8 is more sensitive respect to the refinement ratio. Grid sets containing grid #4 is more sensitive because the uncertainty change is larger as grid refinement ratio increases.
d. CFD’s possible source of error should be more discussed by referring CFD lecture PPT. Modelling & Numerical errors are main parts need to be discussed.
2. Feedback
a. Positive
i. First experience to use ANSYS
ii. Better understanding of CFD process
iii. Visual representation & reinforcement of what he/she learned from the class
iv. Uncertainty analysis was new and interesting v. Pipe flow was a good topic
vi. Learned how to compare simulated data with the actual data vii. Learned about the difference between laminar and turbulent flow viii. Learned about CFD’s accuracy and usefulness
ix. It was a good experience to do hands-on part
b. Negative
i. Required materials are located at the end of the manual ii. Statements in the exercise are not clear
3. Student’s Suggestions
a. Working on a more complex geometry like bended pipe, 3D pipe would be interesting
b. Comparing with 2D channel flow would be interesting
c. More detailed explanation is required for some of the steps like V&V d. Explanation is needed on why k-epsilon is chosen as a turbulent model
