Euler-Bernoulli Beam Deflection Test
Euler-Bernoulli Beam Deflection Test
The literature values of a, b and C are 1, 3 and 4 respectively with zero uncertainty. The
experimental values for a, b and C were found to be 0.9845, 3.1526 and 3.9459 respectively, but
knowing they are integers from (Parkman, T. (2025)) we round them to 1, 3 and 4. A percent
difference analysis was conducted yielding differences of 1.46, 5.09 and 1.35 % for a, b and C
respectively, between the experimental and literature coefficients. The measured values were
also used to calculate the modulus of elasticity for each material yielding 72.14, 97.76, 123.20,
and 203.63 Gpa for aluminum, brass, copper, and 210 steel respectively. The literature moduli of
elasticity for aluminum, brass, copper, and 210 steel are 70, 95, 125, and 210 Gpa respectively. A
precent difference analysis was also conducted on the moduli of elasticities literature and
experimental values resulting in 3.06, 2.91, 3.04, and 3.03% for aluminum, brass, copper, and
210 steel respectively. The differences in experimental and literature values are likely due to
some errors associated with the lab such as dial gauge orientation, beam flatness and dial gauge
stability.
1|Pa ge
OBJECTIVES
The main objective of this experiment is to experimentally determine the constants a, b,
and C by using a three-point bending test apparatus and measuring the deflection of the beam,
height of the beam, width of the beam, material of the beam, and load applied. These values were
measured using a variety of tool which will be listed below.
APPARATUS
Images of each item can be found in the appendix.
2|Pa ge
PROCEDURE
In the experiment we utilized a three-point bending apparatus to experimentally
determine the constants a, b and C. Before our arrival to the lab the beam samples used were
marked at the center and 250mm each way from center. The following steps were then taken:
Basic Procedure:
These are the steps taken for each beam; specific masses and beams will be provided in
each part
Measurement
Sample loading
Dangers: Heavy masses, Loaded objects
1. Attached the beam clamp on the center of the beam sample and tighten with Allen key
2. Placed the beam sample on the three-point bending apparatus aligning the beam clamp
under the dial gauge and the 500mm markings on the rollers at the ends of the apparatus
3. Calibrated the dial gauge to desired rotations by turning calibration screw
4. Placed the desired mass on the mass rack then gently placed the loaded rack onto the
beam clamp
5. Recorded the beam deflection (dial gauge)
Part 1:
1. Performed Sample loading using 510mm x 20mm x 4mm Aluminum beam, 0g of added
mass disks and 8 rotations
2. Performed Sample loading using 510mm x 20mm x 4mm Aluminum beam, 400g of
added mass disks and 8 rotations
3. Performed Sample loading using 510mm x 20mm x 4mm Aluminum beam, 900g of
added mass disks and 8 rotations
4. Performed Sample loading using 510mm x 20mm x 4mm Aluminum beam, 1400g of
added mass disks and 8 rotations
5. Performed Sample loading using 510mm x 20mm x 4mm Aluminum beam, 1900g of
added mass disks and 8 rotations
3|Pa ge
Part 2:
1. Performed Sample loading using 510mm x 20mm x 5mm Aluminum beam, 1900g of
added mass disks and 7 rotations
2. Performed Sample loading using 510mm x 20mm x 6mm Aluminum beam, 1900g of
added mass disks and 7 rotations
3. Performed Sample loading using 510mm x 20mm x 8mm Aluminum beam, 1900g of
added mass disks and 7 rotations
4. Performed Sample loading using 510mm x 20mm x 10mm Aluminum beam, 1900g of
added mass disks and 7 rotations
Note: Additionally reused data from step 5 of part 1
Part 3:
1. Performed Sample loading using 510mm x 20mm x 4mm Brass beam, 900g of added
mass disks and 6 rotations
2. Performed Sample loading using 510mm x 20mm x 4mm Copper beam, 900g of added
mass disks and 6 rotations
3. Performed Sample loading using 510mm x 20mm x 4mm 210 Steel beam, 900g of added
mass disks and 6 rotations
Note: Additionally reused data from step 5 of part 3
4|Pa ge
ANALYSIS
Table 1: Calculated and literature values of constants a, b, and C
Table 2: Calculated and literature values of Young’s Modulus of Aluminum, Brass, Copper, and 210 Steel
1.4
y = 1.0148x + 3.558
R² = 0.9995
1.2
0.8
log10(Force (N))
0.6
0.4
0.2
0
-4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0
-0.2
log10(Deflection(m))
Figure 1: A log vs log plot of Force vs. Deflection with line of best fit within
5|Pa ge
0
-4 -3.5 -3 -2.5 -2 -1.5 -1 -0.5 0
-0.5
-1
log10(Height (m))
-1.5
-2
-3
log10(Deflection(m))
Figure 2: A log vs log plot of Beam height vs. Deflection with line of best fit within
250
200
Youngs Modulus (Gpa)
150
Experimental Values
100 Literature Values
50
0
Aluminium Brass Copper 210 Steel
Material
Figure 3: A bar plot of each materials Young’s Moduli (experimental and literature)
6|Pa ge
The data shows that the coefficients of a, b, and C are approximately 0.9854, 3.1526 and
3.9459 respectively, calculated by equations 4, 7 and 8. Based on (Parkman, T. (2025)) we know
a, b, and C are integer values giving us values 1, 3, and 4 respectively. This result was found
using equation 1 where 𝛿, L, 𝐹, E, w, h, a, b and 𝐶 are the deflection, length, force, Young’s
modulus, width of the beam, height of the beam, and integer constants a, b, and C respectively.
𝐹 𝑎 𝐿3
𝛿= (1)
𝐶𝐸𝑤ℎ𝑏
Utilizing the information found via variable isolatory experiments we concluded a, b, and
C are 1, 3, and 4 respectively allowing the equation 1 to be rewritten as
𝐹𝐿3
𝛿= (10)
4𝐸𝑤ℎ3
This equation can be rearranged to equation 11 and provides a relationship between the
deflection, length, force, width of the beam, height of the beam, and Young’s modulus. This
allows us to experimentally determine the Young’s modulus of any metal giving us 72.14, 97.96,
123.2, and 203.6 Gpa for aluminum, brass, copper and 210 steel respectively. These equations
make sense since a larger force will induce a larger moment producing a larger deflection, more
stiff materials (Higher E values) will tend to deflect less, and the geometry dictates how much
material needs to deflect. The experimental Young’s moduli are slightly different than what the
literature states which are likely due to some errors in the lab which are explored below.
𝐹𝐿3
𝐸= (11)
4𝛿𝑤ℎ3
Analyzing the R-squared values of figures 1 and 2 we can see that the relationship is very
close to the predicted trendline as their values both exceed 0.99 which is the general threshold
for a good quality approximation.
R-Squared Analysis:
Based on an analysis of the R-squared values of figures 1, 2, 4 and 5 we find that both
relationships (height and deflection as well as load and deflection) are likely logarithmic in
nature as their R-squared values of 0.9915 and 0.9995 respectively, are closer to 1. The
relationship between force and deflection could be linear as the R-squared value of 0.9986 is
above the standard 0.99 which means the approximation is an appropriate relationship for the
data but based on our analysis can not be stated as true. For the height versus deflection plot it is
apparent that the relationship is logarithmic as the linear approximation features an R-squared
value of 0.8398, which falls below the acceptable standard.
7|Pa ge
Error analysis:
Dial gauge orientation
The dial gauge used was a SE 100.03 Dial gauge from Gunt Hamburg. This
device features a calibration screw to adjust the depth of the dial however the screw is
used between two vertical parallel plates. This means that when adjusting the number of
rotations, to ensure the dial gauge does not bottom out, the dial gauge changes its angle
relative to perpendicular to the beam sample. The angle was measured via a protractor
and estimated to have a max value of 5° meaning that as the beam deflects downward the
change in distance is not being measured straight down but downward off 5° from
vertical. This would decrease the measured deflection and therefore increase our
coefficients a, b and C and therefore our expected values for the Young’s moduli of each
material. From some preliminary calculation the error associated with this was
determined to be approximately 1% depending on the load, beam material, beam
geometry and angle severity. This was additionally confirmed by readjustment of the dial
gauge in lab and retesting of the sample producing a larger deflection than originally
measured. This error could be minimized by using a vertical slider to prevent the dial
gauge from rotating.
Beam flatness
Each sample used was a rectangular beam which was measured by its length,
width and height. The beams, however, were not considered for their flatness meaning
that if the beam featured a preexisting bow the beams could be stiffer or more ductile.
These preexisting bows were measured to be approximately 1mm producing a difference
of ±4% when loaded with its bow up or bow down respectively (Zdzisław. (2012)). This
change in deformation would increase or decrease (depending on orientation) our
coefficients a, b and C and therefore our expected values for the Young’s moduli of each
material. This makes sense since our aluminum and brass samples were loaded with the
bow up leading to experimental Young’s moduli larger than the literature and the copper
and 210 steel were loaded with the bow down leading to experimental Young’s moduli
smaller than the literature. Uniform heating of each sample would remove any bowing of
the samples and would produce better results.
8|Pa ge
Beam defects
The samples used appeared to be aged by the time we conducted our experiment.
The use of older samples means there could be defects within the samples which can
limit or exaggerate the expected deflection. Some basic defects such as voids, or cracks
can increase the expected deflection. Based on a presumption of mid grade quality
materials and standard crack/void size limitations for each material in can be estimated
that the deflection will increase by 0.5-1% (Matjaž. (2019)). The increased deflection
would decrease our coefficients a, b and C and therefore our expected values for the
Young’s moduli of each material. While this increased deflection may seem insignificant,
each repeated experiment on each sample will only increase this value. The calculation
performed presumed the samples are new enough such that no cracks exists and the
samples have been used a very small number times. Using higher grade materials and
compressive prestressing would remove any internal voids producing more accurate
results.
Dial gauge stability
The dial gauge also featured a significant instability when measuring the
deflection of the beams. This device features a dial which would frequently get stuck
before reading the final value. The dial would get stuck whenever loading or unloading
the sample leading to untrustworthy values. We attempted to rectify this by slight
vibration to the dial gauge to remove any static friction preventing movement but once
the dial was free to move it would oscillate around a small range of 0.1mm. We utilized
the average of where the dial gauge oscillated as our final value. By changing the
deflection by ±0.05mm we discover that the error associated would be approximately
2.5% (Leopold. (2020)). This change in deformation would increase or decrease
(depending on which reading is used) our coefficients a, b and C and therefore our
expected values for the Young’s moduli of each material. This error could be reduced by
lubricating the dial gauge prior to use to ensure that static friction would not influence the
values we see
9|Pa ge
REFERENCES
Pisarek, Zdzisław. (2012). Mechanical Model of Steel-concrete Composite Joint under Sagging
Bending Moment. Selected Scientific Papers - Journal of Civil Engineering. 7. 33-42.
10.2478/v10299-012-0016-z.
Skrinar, Matjaž. (2019). On the Application of the Simplified Crack Model in the Bending, Free
Vibration and Buckling Analysis of Beams with Linear Variation of Widths. Periodica
Polytechnica Civil Engineering. 10.3311/PPci.12698.
Dolzer, D & Leomann, F & Suppan, C & Larour, Patrick & Wagner, Leopold. (2020).
Experimental issues in the instrumented 3 point bending VDA238-100 test. IOP Conference
Series Materials Science and Engineering. 967. 012079. 10.1088/1757-899X/967/1/012079.
Influence of the material microstructural properties on a 3-point bending test Ahmed Maati,
Laurent Tabourot, Pascale Balland, Salim Belaid Mechanics & Industry 21 (5) 518 (2020)
10.1051/meca/2020060
F. Mujika, On the difference between flexural moduli obtained by three-point and four-point
bending tests, Polymer Testing, 2006, 214-220, 10.1016/[Link].2005.10.006.
The Engineering ToolBox (2003). Young’s Modulus of Elasticity – Values for Common Materials.
[Link] November 8, 2025.
Parkman, T. (2025). L16 - Modelling Data [PowerPoint slides]. University of New Brunswick,
ME3701: Mechanical Laboratory I.
210 stainless steel: Properties and Key Applications. Metal Zenith. (n.d.).
[Link]
applications\
Brass Material Properties. Overview of materials for brass. (n.d.).
[Link]
a20e5
10 | P a g e
APPENDIX
100g
Mass Label 100g 400g 500g 900g
(Rack)
11 | P a g e
Table 6: Measured values of when the beam material varies
12 | P a g e
25
20 y = 3430.3x - 0.2536
R² = 0.9986
15
Force (N)
10
0
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007
Deflection(m)
Figure 4: A linear plot of Force vs. Deflection with line of best fit within
0.012
0.01
0.008
Height (m)
0.006
0.004
y = -1.0232x + 0.0093
0.002 R² = 0.8398
0
0 0.001 0.002 0.003 0.004 0.005 0.006 0.007
Deflection(m)
Figure 5: A linear plot of Beam height vs. Deflection with line of best fit within
13 | P a g e
Table 8: Table of equation derivation
𝐹 𝑎 𝐿3
𝛿= (1)
𝐶𝐸𝑤ℎ𝑏
𝐿3
𝑙𝑜𝑔(𝛿) = 𝑎𝑙𝑜𝑔(𝐹) + 𝑙𝑜𝑔 ( ) (2)
𝐶𝐸𝑤ℎ𝑏
1 1 𝐿3
𝑙𝑜𝑔(𝐹) = 𝑙𝑜𝑔(𝛿) − 𝑙𝑜𝑔 ( ) (3)
𝑎 𝑎 𝐶𝐸𝑤ℎ𝑏
1
𝑎= (4)
𝑡𝑟𝑒𝑛𝑑𝑙𝑖𝑛𝑒 𝑠𝑙𝑜𝑝𝑒
𝐹 𝑎 𝐿3
𝑙𝑜𝑔(𝛿) = −𝑏𝑙𝑜𝑔(ℎ) + 𝑙𝑜𝑔 ( ) (5)
𝐶𝐸𝑤
1 1 𝐹 𝑎 𝐿3
𝑙𝑜𝑔(ℎ) = − 𝑙𝑜𝑔(𝛿) + 𝑙𝑜𝑔 ( ) (6)
𝑏 𝑏 𝐶𝐸𝑤
1
𝑏=− (7)
𝑡𝑟𝑒𝑛𝑑𝑙𝑖𝑛𝑒 𝑠𝑙𝑜𝑝𝑒
𝐹 𝑎 𝐿3 −𝑏(𝑡𝑟𝑒𝑛𝑑𝑙𝑖𝑛𝑒 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡)
𝐶= 10 (8)
𝐸𝑤
𝐹 𝑎 𝐿3
𝐸= (9)
𝛿𝐶𝑤ℎ𝑏
𝑥𝑡𝑟𝑢𝑒 − 𝑥𝑒𝑥𝑝𝑒𝑟𝑖𝑚𝑒𝑛𝑡𝑎𝑙
⋅ 100 = %𝐷𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 (12)
𝑥𝑡𝑟𝑢𝑒
14 | P a g e
Calculations:
Aluminium’s Young’s Modulus:
𝐹 ⋅ 𝐿3
𝐸𝐴𝑙 =
𝛿 ⋅ 4 ⋅ 𝑤 ⋅ ℎ3
15 | P a g e
Force exponent (a):
1
𝑎 =
𝑡𝑟𝑒𝑛𝑑𝑙𝑖𝑛𝑒 𝑠𝑙𝑜𝑝𝑒
1
𝑎 = = 0.9854
1.0148
∴ 𝑎 =1
Note: Trendline slope from log (Force) vs. log (Displacement) plot is used in the above calculation
𝐹 𝑎 ⋅ 𝐿3 −𝑏⋅(𝑡𝑟𝑒𝑛𝑑𝑙𝑖𝑛𝑒 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡)
𝐶= 10
𝐸𝐴𝑙 ⋅ 𝑤
(2.00031) ⋅ (9.81) ⋅ (0.500)3 −3⋅(−3.120)
𝐶= 10 = 3.9459
(72.14 ⋅ 106 ) ⋅ (0.02020)
∴𝐶 =4
Note: Trendline slope from log (Height) vs. log (Displacement) plot is used in the above calculation
16 | P a g e
Figure 6: Labelled Schematic diagram of all parts used in lab (Excluding scale)
17 | P a g e
Figure 8: An image of excel table used for calculations
Figure 9: An image of excel table used for calculations showing the formulas used for calculations
Note: Other sections of the sheet did not feature any formulas and were therefore excluded to enhance the image
18 | P a g e
Appendix Question 1:
The assumption of purely elastic deformation presumes that no plastic deformation
occurs, meaning that the preloaded geometry of each beam is the same as the post loaded
geometry. This assumption We can use the three point bending data which keeps the beam
geometry constant as this provides aa direct relationship between the force applied and the stress
within the beam. While we could take some measurements of the beam after loading and verify
that it is the same, we can also prove that verifying that the stress within the beam remains below
the yield stress using equation 13.
3𝐹𝐿
𝜎𝑦 > (13)
2𝑤ℎ2
Since L, w, and h are constants for each beam, the stress within any given beam is
directly proportional to the applied mass or load. We can verify that we remain in the elastic
region for each case by selecting the largest applied load and the smallest beam height since all
other values are constant (or at least presumed constant between bars with negligible
differences).
Table 9: Table of all used materials yield strengths
Material Yield Stress (Mpa)
Aluminum 95
Brass 105
Copper 70
210 Steel 520
3𝐹𝐿
𝜎𝑦𝐴𝑙 >
2𝑤ℎ2
3 ⋅ (2.00031) ⋅ (9.81) ⋅ 0.5
95Mpa >
2 ⋅ 0.02020 ⋅ 0.00412
95Mpa > 43𝑀𝑝𝑎 ✓
3𝐹𝐿
𝜎𝑦𝐵𝑟 >
2𝑤ℎ2
3 ⋅ (1.00031) ⋅ (9.81) ⋅ 0.5
105Mpa >
2 ⋅ 0.02020 ⋅ 0.00412
105Mpa > 21.7𝑀𝑝𝑎 ✓
19 | P a g e
3𝐹𝐿
𝜎𝑦𝐶𝑢 >
2𝑤ℎ2
3 ⋅ (1.00031) ⋅ (9.81) ⋅ 0.5
70Mpa >
2 ⋅ 0.02020 ⋅ 0.00412
70Mpa > 21.7𝑀𝑝𝑎 ✓
3𝐹𝐿
𝜎𝑦𝑆𝑡 >
2𝑤ℎ2
3 ⋅ (1.00031) ⋅ (9.81) ⋅ 0.5
520Mpa >
2 ⋅ 0.02020 ⋅ 0.00412
520Mpa > 21.7𝑀𝑝𝑎 ✓
Therefore, we know that all our testing remained purely elastic
Appendix Question 2:
Using equation 13 as seen in appendix question 1 we know that at the maximum force we
will get exactly the yield stress therefore we can solve the force required using the yield stress.
3𝐹𝑚𝑎𝑥 𝐿
𝜎𝑦 =
2𝑤ℎ2
𝜎𝑦 ⋅ 2 ⋅ 𝑤 ⋅ ℎ2
= 𝐹𝑚𝑎𝑥
3⋅𝐿
95 ⋅ 106 ⋅ 2 ⋅ 0.02 ⋅ 0.0042
= 𝐹𝑚𝑎𝑥
3 ⋅ 0.5
𝐹𝑚𝑎𝑥 = 40.5 𝑁
𝐹𝑚𝑎𝑥 = 𝑚𝑚𝑎𝑥 𝑔
𝐹𝑚𝑎𝑥
= 𝑚𝑚𝑎𝑥
𝑔
40.5
= 𝑚𝑚𝑎𝑥
9.81
𝑚𝑚𝑎𝑥 = 4.077 𝑘𝑔
20 | P a g e