Physics · Article

Models, idealizations, and assumptions

Choose what a model represents, make its simplifications explicit, and check whether its predictions answer the physical question you actually asked.

Previous article: Physical systems, states, boundaries, and environments

Back to Physics

After reading, you can

Distinguish a model, an assumption, an idealization and an initial condition in a simple motion case.
Explain which questions a one-point description of a cart can and cannot answer without its size.
Predict position under a stated constant-velocity assumption while keeping the direction and time interval explicit.
Explain why frictionless motion and constant velocity are different assumptions.
Compare two rules that fit the same supplied readings and identify a useful further check.

Choose a model for a particular question

A model represents selected features of a physical system so that we can describe, explain or predict something about it. A diagram, a physical replica, an equation and a computer simulation can all be models. The first choice is the question: the position of a cart, the time its front reaches a line, or whether it will turn over. Those questions need different information.

A drawing that keeps only a cart's tracked point can help organise its motion along a straight track. It leaves out the cart's shape. That omission may suit a position question, but a question about clearance, contact or tipping needs additional features. Write down the model's purpose before judging its detail.

Separate assumptions from inputs and starting conditions

An assumption states something the model takes as given for the question being investigated. An idealization is a deliberate simplification, such as treating an extended object as a point or setting a small resistance to zero. Idealizations are assumptions about the representation; other assumptions can specify the situation without removing detail.

In our cart case, the tracked mark starts at 0.40 m when the clock reads zero. That is an initial condition. The chosen value 0.25 m/s is an input, and keeping that velocity unchanged towards increasing positions throughout 0–4 s is an assumption. These numbers are stipulated teaching values. In a real investigation, record whether an input was measured, fitted, estimated or chosen.

Know what a point representation leaves out

A point-particle model represents an object's translational motion without describing its extent or internal shape. Introductory motion teaching often uses this approximation where the object's dimensions are small compared with the motion distances of interest. Suitability also depends on the requested quantity; a size comparison alone does not answer every modelling question.

A tracked-point model can report a cart midpoint's position while leaving out its front and rear. To predict when the front crosses a nearby line, keep the cart's length and orientation as well. To describe rotation, deformation or contact geometry, add the corresponding features. Calling an object a particle in this model does not claim that the real cart has zero size.

Give a motion rule its direction and interval

For the stipulated tracked mark, use x = 0.40 m + (0.25 m/s) × t during 0–4 s. Here t is elapsed time from the stated start, and x is position along the same straight track axis. The positive coefficient means motion towards increasing positions. Constant speed without a fixed direction would not supply this whole rule.

This rule describes motion; it does not identify the mechanism maintaining it. For a Newtonian point-particle model, constant velocity follows from zero net external force in an inertial reference frame. Forces may balance rather than vanish individually. For an ordinary short track experiment, treating the laboratory frame as approximately inertial is itself a modelling choice.

Make omitted resistance a separate assumption

Frictionless is an idealization in which the model omits the relevant friction force. A visually smooth surface does not establish that friction is absent. Friction depends on the contacting systems, and simple sliding-friction equations are approximate empirical descriptions. A sliding-friction model is not automatically the right account of a cart rolling on wheels.

Ignoring friction alone does not guarantee constant velocity: another unbalanced force could change the motion. Constant velocity also does not prove zero friction, because other forces may balance it. To decide whether resistance can be neglected, compare the consequences of including or omitting it for the required interval and accuracy, using appropriate measurements or a justified estimate.

Use a fit as a starting point for checking

Fitting chooses a model's adjustable values to agree with existing data. Agreement at those same points is useful, but it is not a separate prediction check. Different rules can agree with a short record and disagree elsewhere. Prediction within the sampled range and prediction outside it both need scrutiny; going beyond the range is called extrapolation.

Keep the rule and its fitted values fixed when checking additional observations. Compare the same tracked point, time reference and units, and consider measurement limitations. A disagreement can involve the model, its inputs, the records or the measurement process. A successful check supports the tested use; it does not prove every assumption or establish unlimited validity.

Check the calculation and the physical representation

For a calculation, check the arithmetic, units and starting conditions. For a simulation, also check that the program implements the chosen rule and inputs correctly; software testing is part of making data processing reliable. A correctly implemented rule still needs evidence that it suits the physical system and question.

When revising, state what changed and why: include the cart's length for a front-crossing question, restrict a constant-velocity rule to the interval supported, or investigate resistance before changing the motion model. Preserve the original record. More detail is useful when it addresses an important missing feature, rather than merely making the representation look more realistic.

Choose, state, predict and check

1. Chosen question
Identify the tracked point, event or physical feature the model needs to describe.
2. Stated assumptions
Record the retained geometry, direction, motion interval, inputs and omitted effects.
3. Conditional prediction
Use the chosen rule and inputs to calculate the requested position or crossing time.
4. New check
Compare suitable additional observations with the fixed prediction, consider measurement limitations, and investigate any disagreement.

Keep the chosen question and assumptions attached to a prediction. Additional observations help check whether the model suits its purpose.

Worked examples

Predict a position under stated assumptions

A tracked mark starts at 0.40 m at t = 0 s. A teaching model assumes straight motion towards increasing positions at an unchanged 0.25 m/s during 0–4 s. What position does it predict at 3 s?

Keep the given starting position rather than replacing it with zero.
The predicted position change is 0.25 m/s × 3 s = 0.75 m.
Add the starting position: 0.40 m + 0.75 m = 1.15 m.

The model predicts 1.15 m at 3 s. The arithmetic is exact for the stipulated inputs; it does not show that a real cart has maintained this velocity or justify using the rule after 4 s.

Keep extent when the question depends on it

A rigid cart is 0.20 m long and aligned with a straight track, with its front towards increasing positions. At t = 0 s its midpoint is at 0.80 m. The whole cart translates at an unchanged 0.40 m/s without rotating. When will its front reach the line at 1.00 m, and when will its midpoint reach that line?

The front is half the cart length ahead of its midpoint: 0.80 m + 0.10 m = 0.90 m.
The front has 0.10 m left to travel. At the stipulated velocity its crossing time is 0.10 m ÷ 0.40 m/s = 0.25 s.
The midpoint has 0.20 m left to travel, so its crossing time is 0.20 m ÷ 0.40 m/s = 0.50 s.

The front crosses at 0.25 s and the midpoint at 0.50 s. A point-only model reporting the midpoint's crossing would answer the wrong event question if it were used unchanged for the front. The length, alignment, rigidity and no-rotation assumptions supply the relationship.

Two rules can fit the same readings

A supplied position record contains only 0.40 m at 0 s and 0.65 m at 1 s. Compare rule A, x_A = 0.40 m + (0.25 m/s) × t, with rule B, x_B = 0.40 m + (0.20 m/s) × t + (0.05 m/s²) × t². Here t² means t multiplied by t. Do these two recorded points select one rule? What does each predict at 2 s?

At 0 s the time terms vanish, so both rules give 0.40 m.
At 1 s, A gives 0.40 m + 0.25 m = 0.65 m; B gives 0.40 m + 0.20 m + 0.05 m = 0.65 m.
At 2 s, A gives 0.40 m + 0.50 m = 0.90 m; B gives 0.40 m + 0.40 m + 0.20 m = 1.00 m.

The two supplied points alone do not select one rule. An additional suitable observation could help distinguish these candidates. Neither formula is established as a real force mechanism or a law of motion by this invented fit.

Try it yourself

Classify these in the cart case: represent the cart by one tracked mark; start that mark at 0.40 m at t = 0 s; choose 0.25 m/s; keep that velocity unchanged towards increasing positions throughout 0–4 s.
Show the answer and reasoning
Idealization of the representation; initial condition; chosen model input; motion assumption with direction and interval.
The one-point representation deliberately omits extent. The starting coordinate and chosen velocity value are inputs to the rule, while its continued use requires the stated motion assumption. Inputs may have different evidential origins in a real investigation.
Using the stated constant-velocity model, what position is predicted at 4 s? Would 1.00 m answer that position question?
Show the answer and reasoning
1.40 m. The 1.00 m value is the position change, not the final position.
Calculate 0.25 m/s × 4 s = 1.00 m, then add the 0.40 m starting coordinate. The chosen origin and starting condition remain part of the model.
The formula produces a number at 8 s, but the stated motion assumption covers only 0–4 s. Does the arithmetic alone justify the physical prediction at 8 s?
Show the answer and reasoning
No. It extends the calculation beyond the stated use.
The expression gives 0.40 m + (0.25 m/s) × 8 s = 2.40 m. That value is conditional on continuing the rule, which has not been justified for 8 s. Seek appropriate evidence or another model.
A model neglects friction. Has it therefore established constant velocity? Does observed constant velocity establish that no forces act?
Show the answer and reasoning
Neither conclusion follows.
An unbalanced force other than friction can change velocity. Conversely, individual forces can balance to give zero net force. State the complete relevant assumptions instead of replacing them with the word frictionless.
A point representation correctly reports a cart midpoint's position. Can it by itself determine whether the finite cart will fit through a narrow opening?
Show the answer and reasoning
No. The clearance question also needs relevant dimensions, shape and orientation.
A representation useful for one position quantity can omit information essential to a different question. Extend the representation to include that information.
Rules A and B in the worked fit both match the supplied records at 0 s and 1 s. At 2 s a separately supplied comparison interval is 0.98–1.02 m. Which rule's prediction lies within it, and what does that establish?
Show the answer and reasoning
B's 1.00 m lies within the interval; A's 0.90 m does not.
The interval is stipulated teaching information with no inferred statistical meaning. B is compatible with this particular check and A is not. This does not prove B's mechanism or its behaviour at other times; inspect the relevant records, measurement information and assumptions.

What to carry forward

Choose the physical question before selecting the model's detail.
Record the idealizations, chosen inputs, initial conditions and assumptions separately.
A point representation leaves out extent; retain size and orientation when the requested quantity depends on them.
Constant velocity requires a direction and a stated use interval; omitting friction is a different assumption.
Check new observations as well as the data used for fitting, and retain the limits of any agreement.
Verify calculations or program implementation, then assess whether the physical representation suits the question.

References

Sources for the concepts and teaching approaches in this article.

Physics · 1.2 The Scientific MethodsPaul Peter Urone and Roger Hinrichs · OpenStaxTextbook chapteropenstax.orgRead source
Physics Part I · Chapter Two: Motion in a Straight LineNational Council of Educational Research and Training (NCERT), IndiaTextbook chapter · PDFncert.nic.inRead source
University Physics Volume 1William Moebs, Samuel J. Ling and Jeff Sanny · OpenStax / Rice UniversityTextbook chapterUniversity Physics Volume 1 · 5.2 Newton's First Law University Physics Volume 1 · 6.2 Frictionopenstax.org
NIST/SEMATECH e-Handbook of Statistical Methods · 5.5.9.9.8 Motivation: How do we Use the Model Beyond the Data Domain?National Institute of Standards and Technology (NIST) / SEMATECHReference guideitl.nist.govRead source
Data analytics and modellingNational Physical Laboratory (NPL), United KingdomResearch project overviewnpl.co.ukRead source