ENGG2500 Chap.6 Dimensional Analysis and Similarity
Dimensional Analysis and Similarity
Dimensional analysis reorganises a physically selected variable set into independent dimensionless groups. Buckingham Pi predicts the number of groups and provides an exponent method for building them. Model testing then requires geometric, kinematic and dynamic similarity.
Reynolds number is central when inertia and viscosity compete, but other force ratios must be matched when gravity, compressibility or surface tension matters. The variable list is a physical claim, and the final groups must be dimensionless, independent and interpretable as response measures or ratios of competing effects.
What this chapter covers
- 01
Variable selection
- 02
Base dimensions and rank
- 03
Buckingham Pi groups
- 04
Repeating variables
- 05
Geometric and dynamic similarity
- 06
Reynolds scaling and distorted models
Worked example · free
Reynolds scaling of a model
- 1Set model and prototype Reynolds numbers equal. The marking scheme used here is not an official assessment scheme.
- 1Cancel the common kinematic viscosity.
- 1Use model length equal to prototype length divided by five.
- 1Model speed must be 15 m/s.
Key terms
- Pi group
- A Pi group is an independent dimensionless product of governing variables.
- Repeating variables
- Repeating variables form a dimensionally independent basis used to construct groups.
- Dynamic similarity
- Dynamic similarity matches the relevant force ratios between model and prototype.
- Reynolds number
- Reynolds number compares inertia with viscous effects.
Dimensional Analysis and Similarity FAQ
How many Pi groups should I obtain?
If n variables use r dimensionally independent base dimensions, Buckingham Pi gives n minus r independent groups. Verify the dimensional rank rather than merely counting symbols, then expand units to confirm that every completed group is dimensionless and that none can be formed from the others.
How do I choose repeating variables?
Choose variables that collectively contain every base dimension, are dimensionally independent and do not already make a dimensionless product. A characteristic density, speed and length often works for inertial flows, but the physically correct set depends on the variables chosen for the phenomenon.
What if two similarity requirements conflict?
Exact dynamic similarity may be impossible with the available scale and fluid because different force ratios require different speeds. State which mechanism dominates, match its group, calculate the mismatch in the secondary group and explain how that scale effect limits interpretation of the experiment.
Why is variable selection so important?
Buckingham Pi can only reorganise variables that were included. If gravity, viscosity, compressibility or surface tension affects the mechanism but is omitted, no exponent calculation can recover it. Start from a force or transport argument and justify the response variable, geometric scales, fluid properties and operating quantities.
Can two different Pi-group sets both be correct?
Yes. Different valid repeating variables can produce groups that look different but are algebraically related. They describe the same dimensionless relationship if each set is independent and complete. Rearrange the groups to a familiar coefficient or force ratio before comparing results from two derivations.
What is geometric similarity?
Geometric similarity means corresponding lengths have a common scale factor, so angles and shape ratios match. It is necessary but not sufficient for dynamic similarity. A geometrically accurate model can still predict the wrong behaviour if Reynolds, Froude, Mach or another relevant force ratio differs from the prototype.
How should dimensionless experimental data be plotted?
Place a dimensionless response, such as force coefficient, against the controlling dimensionless input groups. Use separate curves or symbols for a second group such as relative roughness. Successful collapse of different sizes or fluids onto one relation is evidence that the selected groups capture the scaling behaviour.
What makes a Pi group independent?
An independent group cannot be formed by multiplying powers of the other groups. Independence matters because redundant groups add no new experimental coordinate and can hide a missing influence. Compare exponent vectors or solve the dimensional matrix rank to confirm that the final set spans the required null space.
Exam move
Begin each derivation with a sentence naming the response and the physical influences that could change it. Write every variable in mass, length and time dimensions, determine dimensional rank and predict the number of Pi groups before choosing repeating variables. Select repeaters that contain all base dimensions, are independent and do not already make a dimensionless product.
Solve one exponent system per remaining variable, verify every group by expanding units and compare the final count with n minus r. Rearrange groups into recognisable quantities such as Reynolds number or a force coefficient and explain the competing effects. For model problems, write model-to-prototype ratios before inserting a geometric scale.
Match the dimensionless groups demanded by the dominant physics, solve for speed or property requirements and test whether they are practical. If two requirements conflict, quantify the unmatched group and label the model distorted rather than forcing an impossible exact match. Finish by stating what the experimental axes should be.