Stacking Bricks Model
The Simple Model
Using a meter stick, measure the height of a variety of stacks of bricks to the nearest 0.1 cm
and record in the table below. Stack the bricks in their lowest energy positions!
Number of bricks Height of stack, cm
Using a spreadsheet, derive a linear mathematical model for the data and record the model in
terms of the variables measured (not x and y) and r-squared below.
Define slope in terms of the model variables. What does the slope represent?
What should the value of the y-intercept be for the model?
What does the y-intercept represent? What could cause the value?
Predict the height of 100 bricks. Explain/show how you determined this value.
NASA Langley Modeling & Simulation Workshop – Sinex (2014) 1
Field Work to Verify the Model
Find a real brick wall and repeat your measurements for the same number of bricks.
Number of bricks Height of stack, cm
Sketch how you measured a stack of bricks in the wall. Record any observations.
Derive a linear mathematical model for the data and record the model in terms of the variables
measured and r-squared below.
Define slope in terms of the model variables. What does the slope represent?
NASA Langley Modeling & Simulation Workshop – Sinex (2014) 2
Deriving a Multivariable Model
Measure the height of five different bricks in the wall and determine the average height of a
brick.
Brick Height, cm
#1
#2
#3
#4
#5
average
σ
%CV
Calculate the sample standard deviation, σ, using =STDEV() and the percent coefficient of
variation, %CV = (sample standard deviation/mean) x 100. The smaller the %CV value is the
less variation in the measurement.
Measure the thickness of the mortar (white cement between the bricks) at five different places
in the wall and determine the average thickness of the mortar.
Mortar Thickness, cm
Place 1
Place 2
Place 3
Place 4
Place 5
average
σ
%CV
Which has more variation – brick height or mortar thickness? Why from a statistical point?
Why from a physical point? Think about the bricks and the brick laying process.
NASA Langley Modeling & Simulation Workshop – Sinex (2014) 3
Compare your two models. Explain any differences between the two models. Easiest way to
do this is to plot both sets of data on the same graph.
Write generic equations in terms of the possible experimental variables for the two different
mathematical models.
List of variables Simple stacking model:
H = total height, cm
n = number of bricks in stack
b = height of a brick, cm Realistic Model from wall:
t = thickness of mortar, cm
Predict the height of 100 bricks in a real wall. Explain how you determined.
Investigating with a Brick Height Simulation
Now let’s go to the Brick Height Simulator to be sure we understand the models we developed.
The interactive Excel spreadsheet can be obtained at:
http://academic.pgcc.edu/~ssinex/excelets/brick_height.xls
Go to the “brick model 1” tab on the spreadsheet. Is your model from the actual
measurements as good as the one on the spreadsheet? Why or why not?
Are your bricks a consistent height? Explain.
Now go the “brick model 2” tab. What is the height of a brick plus the mortar?
NASA Langley Modeling & Simulation Workshop – Sinex (2014) 4
What is the cause of the difference between the two lines (models)?
Of the two models, which one has the better chances of being the more realistic model?
The error between the two models is a proportional systematic error. The error grows as the
number of bricks increases since as a brick is added so is a layer of mortar. The realistic model
from the wall is the better model; hence, the simple stacking of brick model suffers from a
negative proportional systematic error (lower slope) due to the lack of considering the mortar
in between bricks.
From your replicate measurements of brick height and mortar thickness on page 3, which
measurement has more variation? This is random variation due to the manufacturing of the
bricks and the laying of the mortar layer.
Now go the “bricks model 2 error’ tab. Here we can add random error to the mortar thickness
and observe what happens to the model. Set a mid-range of random error and then hit the F9
key multiple times to see what happens. Summarize how the model is influenced.
There is random error or variation contributed from both the brick height variation due to
manufacturing and the thickness of the mortar between bricks. This causes the slope and y-
intercept to randomly vary and more scatter in the data points, which lowers the r-squared
value. Measurement error could also by contributing to this as well. This is why your models
have small nonzero y-intercepts.
Try the “brick assessment” tab to be sure you understand the influence of variation.
NASA Langley Modeling & Simulation Workshop – Sinex (2014) 5
Notes to Teachers
The activity shows how a simple model can be derived using spreadsheets and then enhanced
by verifying and discovering the mortar layer and how it influences the model – good science
process! Random variation is also explored and how it influences the models.
Bricks are stacked in there lower energy position – largest surface area facing down.
Materials needed for the activity: at least 10 bricks of the same style, meter sticks, and 15-cm
rulers. Access to an exterior finished brick wall of any type is also needed. A picture with
something used as a scale in it would also work. Students should have experience with
spreadsheets – entering data, graphing, and linear regressions.
Students will derive two different mathematical models:
1. Simple model from stacking the bricks in the classroom: H = bn slope = b
2. More realistic model from measuring on the brick wall: H = (b + t)n slope = (b + t)
The variables are defined as H is total height of bricks, n is number of bricks in a stack, b is
average brick height, and t is the average thickness of the mortar layer between bricks in the
wall.
Variation of brick height during the manufacturing process, variation of the thickness of the
mortar layer, and measurement variation will cause the models to have a small nonzero y-
intercept. Both models should have high r-squared values which shows very good linear fit.
Beware that different styles of bricks have different brick heights (5.4 - 7.3 cm) and that mortar
layers can vary from 0.9 to 1.3 cm. In the US, bricks must meet ASTM standards, which produce
uniform size of the various styles in the manufacturing process.
The Brick Height Simulator (http://academic.pgcc.edu/~ssinex/excelets/brick_height.xls) will
allow students to investigate the models and be sure that they develop an understanding of the
multivariable model and the influence of random variation.
NASA Langley Modeling & Simulation Workshop – Sinex (2014) 6
