Tensions in Practice: Coefficient-free reasoning in tension with absolute prediction¶
Physical model · length and acceleration
Assume a model family in which a period T depends only on length L and acceleration g, with every dimensionless condition held fixed. Dimensional reasoning permits T=C√(L/g), leaving C unknown. At fixed g, multiplying length by four doubles the period within this model. Predicting seconds needs one more anchor: a value for C or a reference period.
Use scaling without calibration
Compare relative periods before obtaining an absolute reference.
Predict an absolute period
Attach a measured or derived factor to the dimensional form.
Why these aims pull against each other
The units constrain exponents but do not supply the dimensionless factor. Ratios can cancel that factor; absolute predictions cannot.
Choose an arrangement to see what changes and what remains difficult.
Finite illustrative comparisons. Text states carry the meaning; color is not a measured score or universal preference.
What this choice protects
What it costs
When it fits
Compare the arrangements
Ratios only
Use T/T₀=√(L/L₀) at fixed g and within the same declared model regime.
| Length ratio | Period ratio | Period | |
|---|---|---|---|
| Base | 1 | 1 | Not known |
| Longer | 4 | 2 | Not known |
| Longest | 9 | 3 | Not known |
- What it protects
- The relative predictions need no reference period in seconds.
- What it costs
- The table cannot say how many seconds any period lasts.
- When it fits
- Relative comparisons answer the question and the model’s complete-variable and fixed-regime assumptions are defensible.
Illustration note: The model family and ratios are editorial assumptions. Dimensional consistency alone does not prove that a real system follows this family.
Reference period
Stipulate an ideal reference reading T₀=2 seconds at L₀; apply the same length ratios and unchanged g.
| Length ratio | Period ratio | Period | |
|---|---|---|---|
| Base | 1 | 1 | 2 s |
| Longer | 4 | 2 | 4 s |
| Longest | 9 | 3 | 6 s |
- What it protects
- The same relative form now yields explicit periods of 2, 4 and 6 seconds.
- What it costs
- A trustworthy reference must be obtained and its uncertainty and applicability maintained; a regime change can invalidate the reused factor.
- When it fits
- Absolute predictions are needed and a valid reference or independent dynamical derivation is available.
Illustration note: The 2-second reference is invented for this finite example, not an empirical reading. It is treated as exact to isolate the missing-factor issue.
What this illustration does—and does not—establish
Dimensional Analysis: Dimensional reduction power vs functional form recovery supplies the unfilled prefactor; its explicit physical-correctness limit bounds the ratio/calibration illustration.
- No claim is made that length and acceleration exhaust the variables of every real oscillator.
- Angles, damping or other omitted dimensionless conditions can change the factor or model form.
- Calibration cannot repair an incorrect functional model merely by supplying one number.
Source entries
Dimensional Analysis
Dimensional Analysis: Dimensional reduction power vs functional form recovery supplies the conflict examined here.
Dimensional reduction power vs functional form recovery
Dimensional analysis gives form up to a dimensionless prefactor; experiment determines the prefactor. The method's fundamental limitation is that it prunes the space of possible answers but cannot uniquely select one.
What It Is Not
- Not a guarantee of physical correctness. A dimensionally consistent equation can still be physically wrong; dimensional homogeneity is necessary, not sufficient. A dimensionally correct model may omit a crucial physical process entirely.