In the Laboratory

Polymer Mechanical Properties via a New Laboratory

Tensile Tester
T. Carter Gilmer* and Matthew Williams
Department of Chemistry, University of Michigan-Dearborn, 4901 Evergreen Road, Dearborn, MI 48128-1491

Polymers, or macro- Table 1. Categories of Polymers

molecules, are gaining in-
terest in the technical Natural Synthetic Bio/Synthetic, or
and scientific communi- (Bio) Hybrids
ties. There is an abun- Organic Organic Inorganic Organic
dance of new and useful Proteins Fibers SiO2 – based glass Vulcanized (natural rubber) tires
polymeric structures that hair Nylon – carpets
generate corresponding silk Antron – upholstery
wool Orlon – wood substitute
new, interesting, and po- Kevlar – nylon based composite,
tentially useful (mechani- bullet proof vests
cal) properties. Table 1 Nucleic Acids Plastics Silicones Genetically engineered DNA
shows a way of categoriz- (DNA/RNA) Plexiglas
ing natural and syn- Contact lens
thetic polymers (1). The Corian – marble substitute
subcategorization that Polysaccharides Elastomers
will be emphasized di- wood Rubber Bands
starch Spandex – swimwear
vides solid material ori-
ented polymers into:
• fibers (low elongation, high stiffness and modulus) as tensile strength)
• plastics (intermediate elongation, generally inter- 4. Yield (applies primarily to plastics but also to rub-
mediate stiffness) bers, where stress/strain slope equals zero; polymer
• elastomers or rubbers (high elongation, low modulus) chains are disentangling and rearranging causing
no additional stress with increasing strain), deter-
Examples of natural fibers are wool, silk, and wood. mined at a specific stress or strain point
Common synthetic fibers are Nylon® (polyamides) and
5. Toughness or energy-to-break, area under the
Rayon®. Rubbers are more readily identifiable. Natural
stress/strain curve to the break point (material
rubber (polyisoprene) comes from trees. Rubber bands
must be strained to failure)
and tires are common examples of (hybrid) synthetic
elastomers. But plastics are all synthetic, originating The following experiment on tensile properties de-
with Bakelite discovered by Leo Bakeland in the early scribes a simple laboratory version of a common sophis-
1900s. Synthetic plastics are ubiquitous now. Eyeglass ticated tensile machine, namely, Instron’s Tensile Testers.
lenses are often made of polycarbonates. Trash bags are A polymeric material of known dimensions (width and
often polyethylene. Cookware is coated with a nonstick, thickness primarily and length secondarily) is attached
high surface energy Teflon® or tetrafluoropolyethylene. to and suspended from a laboratory clamp. Attached to
Many demands are placed on each specific polymer, the bottom of the sample is a bucket with a handle, into
polymer blend, or polymeric composite. Essentially all which precise weights can be placed. As individual
materials, be they substitutes for brick, wood, glass, or weights are added to the bucket, a specific load is ap-
other structurally demanding materials, must have and plied, which upon using the known cross-sectional area
reasonably maintain specific mechanical properties. One of the sample is converted to stress (force/unit area). By
way of measuring a range of some mechanical proper- measuring the initial length of the sample and its in-
ties is to perform tensile and/or compression testing. cremental increase with each new load, strain is also
Discussions here will refer to determination of ten- recorded. Upon completing the process of adding weights
sile properties in a uniaxial mode. That is, stress is ap- and measuring each new length, all necessary data for
plied along one axis, as when you stretch a rubber band a standard stress/strain curve are generated. Some
by extending your hands. Figure 1 shows three repre- samples will break with the maximum weight applied;
sentative stress/strain curves of a typical fiber, a plastic others may not break under these conditions.
and an elastomer, which have undergone tensile testing During the tensile testing basically three events oc-
(2, 3). As described in the attached Tensile Properties cur that can lead to fracture or breakage:
Experiment, five common parameters are obtained from • disentanglement and stretching of the polymer:
each curve: random coils stretch and align in a somewhat par-
1. Tensile strength (in lb/sq in. or MPa) (force/area) allel fashion, and crystalline polymers in a lamel-
2. Elongation-to-break (% increase in length) lar arrangement unalign and go towards a
3. Elastic modulus or Young’s modulus or modulus stretched-out, somewhat linear configuration
(measure of stiffness, or stress/strain—same units • breakage of covalent bonds break under the applied
stress (see Fig. 2)
• chain slippage (where secondary bonds, e.g. dipole–
*Corresponding author. Present address: Department of dipole interactions, are broken; see Fig. 2)
Chemistry, Bowling Green State University, Bowling Green, OH
43403-0213. In tensile testing it is critical to use uniform samples

1062 Journal of Chemical Education • Vol. 73 No. 11 November 1996


Figure 1. Stress–strain behavior of three types of polymeric
materials.
Figure 2. Schematic representation of fracture mechanisms in poly-
meric materials: (a) bond breakage; (b) chain slippage.
Figure 3. Variation of the stress–strain behavior of poly(methyl-
methacrylate) with temperature.
Vol. 73 No. 11 November 1996 • Journal of Chemical Education
In the Laboratory
with minimal surface defects or flaws. If a sample is non-
uniform in terms of density or cross-sectional area, the
least dense sections or the thinnest will have a lower
effective cross-sectional area, which under a given load
(weight) will experience the most stress. As cross-sec-
tional area decreases, stress increases, assuming the load
is the same. So the thinner or less dense sections, or sec-
tions with flaws, become the locations most likely to fail
first. In effect these locations are the weak links in the
chain. Accordingly, tensile measurements such as
strength, elongation-to-break, and toughness will be bi-
ased towards the low side if defects and/or thin sections
exist. To get the most accurate and precise tensile data
the sample must be uniform in dimensions and contain
minimal flaws, especially surface flaws.
All of the discussion so far has implied tensile mea-
surements done at room temperature or 20 °C. Often
polymers exhibit tremendous change in mechanical prop-
erties with temperature (4). This is readily illustrated
in Figure 3 for poly(methyl methacrylate) (PMMA). As
shown in Figure 3, PMMA appears to be a brittle plas-
tic (resembling a fiber) at room temperature or 277 K.
PMMA’s modulus steadily decreases as temperature in-
creases, with corresponding changes in other mechani-
cal properties. At 333 K, the stress/strain curve shape
for PMMA resembles that of an elastomer.
On a macroscopic scale tensile data are quite infor-
mative as they reflect the collective effect of many poly-
mer chains. With more in-depth analysis they can be use-
ful in obtaining microscopic information as well. For ex-
ample, polymers that exhibit extensive H-bonding gen-
erally are stiffer (higher modulus) than those with less
H-bonding. Conformation of polymers (tacticity, for ex-
ample) also affects their morphology (shape), which af-
fects entanglements, density, and in turn mechanical
properties.
Many polymers are solids. Mechanical properties of
all solids are definable to some extent and extremely use-
ful information can be obtained from those studies.
Again, one difficulty in obtaining meaningful data is as-
suring that an adequate sample is prepared that is uni-
form and free of significant surface defects.
Apparatus and Experimental Procedure
Tensile measurements are readily determined by
individual students in a small laboratory. Only common
and quite inexpensive tools and devices are needed. Be-
ginning with a sample suspended from a laboratory rack
and attached to a load-carrying bucket, stress versus
strain curves are determined as described below.
Procedure for Tensile Properties Determinations
(of a Fiber, a Plastic, and an Elastomer)
Materials
• Ninety 32.0-g bolts (or an equivalent constant-mass
material)
• 1 light-weight basket with removable handle
(model is 129 g with aluminum mesh screen)
• Regular typing paper, known thickness; 4.5 mil
(0.001 in.) is fine
• Rubber bands (measure thickness and width)
• 1 elastic strap (about 8 in. long) with metal hooks
at ends (often called bike strap or bungy cord)
• 1 roll of polyethylene film (0.8 mil thickness)
• Scotch tape
• 1 tape measure (or meter stick)
• Micrometer
1063
In the Laboratory

Figure 4. Stress/strain plot for polyethylene.

Figure 5. Modulus of polyethylene.

Procedure Results
1. Attach one hook of the elastic strap to a lab clamp. As the procedure above describes, four types of poly-
2. Adjust clamp location so that with strap extended meric materials were chosen for demonstration purposes:
and attached to the basket with its handle, the bot- a fiber (paper), a plastic (polyethylene), an elastomer
tom of the basket is about 10 in. from bench top. (rubber band) and a composite (bike strap). Using this
3. Measure length of the strap excluding metal hooks. laboratory apparatus, a stress/strain curve for a commer-
4. Incrementally add specific loads (maybe 5 bolts) cially available polyethylene (PE) was generated (Fig. 4).
(weights) to the basket. Subsequently, this curve was used to calculate:
5. After each additional load measure the new length
and record the total weight. • tensile strength 11.9 MPa
6. Continue steps 3–5 until the material breaks or the • elastic modulus 41 MPa
basket is full with all 90 bolts or 2880 g. • % elongation-to-break (%E) 366%
7. Using paper strips 1 in. wide and 12 in. long, rein- • toughness (area under
34 MPa
force the top and bottom 4 in. with Scotch tape. stress/strain curve)
8. Wrap the top taped portion of the paper around the Figure 5 shows an expanded scale (x-axis) to look
support (lab clamp) and secure it by reinforcing more closely at the linear portion of the stress/strain
that portion with more tape, making a loop around curve and more accurately determine modulus.
the clamp. A similar stress/strain curve is shown for a rubber
9. Repeat step 8 for the bottom portion and slip this band in Figure 6. Here, the percent elongation-to-break
loop through the handle of the basket. (432%) is higher than for the PE, and the modulus and
10. Measure the length of the paper (approximately 4
in.), excluding the portions that have been rein- Table 2. Tensile Properties of Polyethylenesa
forced by Scotch tape. Tensile Strength Modulus Elongation-to-break
11. Repeat steps 4–6. (MPa) (MPa) (%)
12. Using the polyethylene, which has been cut with a
LDPEb 6.9 – 17.2 138 – 310 100 – 700
razor blade or scissors to 1/2 in. (or any other
known thickness from about 1/16 to 1 in.) and 12 LLDPEb 14 – 21 137 – 186 200 – 1200
in. long, attach top and bottom portions to clamp HDPEb 18.6 – 30.3 — 100 – 1000
and reinforce with transparent tape as in step 7. UHMWPEb 19.9 – 41.4 110,000 300
13. Continue from step 12 with steps 8–11.
14. Stress–strain data can be collected for all other ma- LMDPEb 14 – 24 — 200 – 1200
terials (such as the rubber bands) in a way similar "LLDPE"c 12 41 360
to that described in steps 7–11. a
LDPE= low-density polyethylene (PE); LLDPE = linear low density
15. Plot load (force or weight added) versus strain for PE; HDPE = high density PE; UHMWPE = ultra-high molecular weight PE;
each material. LMDPE = linear medium density PE.
16. Plot stress (force/cross-sectional area) versus strain b
Values from ref 6.
for each material. Strain = (new length – initial c
Values via Gilmer’s Laboratory Tensile Device. “LLPDE” is Brute
length)/initial length. Kitchen Bags™ – a blend of LLDPE (linear low-density PE); also contains TiO 2.
17. From plots in 16 above, determine tensile strength
(peak stress obtained), elastic modulus (slope of lin-
Table 3. Precision of Weight of a Set of "Constant" Weight
ear portion of stress/strain curve), elongation-to-
Objects (Bolts)
break (strain at break × 100), and toughness (or
energy-to-break), area under the stress/strain curve. Number weighed 20
Sample must break for toughness measurement. Range 22.24 – 22.40 g
Tensile strength in MPa (Pa = N/m 2); modulus in Mean 22.32 g
MPa; toughness in MPa (or kJ/m3). 1 MPa = 145 psi;
1 N = 0.2248 ft-lb; N = (meter) (kg)/s2; F (force) = Standard Deviation 0.046 g
ma (mass × acceleration). % Standard Deviation 0.21%

1064 Journal of Chemical Education • Vol. 73 No. 11 November 1996


Figure 6. Stress vs. strain of rubber band.
tensile strength of the rubber are much lower—1.52 MPa
and 4.17 MPa, respectively. When PE and the rubber
bands were run in duplicate tests, the results were very
reproducible, although in some cases %E and tensile
strength varied. This variation is normal in tensile test-
ing, since failure (breakage) is so dependent on flaws in
the test specimen. Samples with more flaws or larger
flaws fail sooner. Thus, reproducibility is most dependent
on sample uniformity and sample preparation.
Since there are several grades of PE, we decided to
compare mechanical data for our “Brute Kitchen Bags” PE
generated by our apparatus to other values in the litera-
ture. As shown in Table 2, tensile properties for various
grades of PE vary significantly between grades and show
wide ranges in values even within a specific type or grade.
In measuring tensile properties of a composite
sample (in this case a bike strap or bungy cord), the
stress-strain curve shows two separate response regions
for the two components of the composite. Those two com-
ponents are (i) the interior elastic rubber and (ii) the ex-
terior fibrous woven cord. See Figure 7. As stress is ap-
plied the rubber matrix responds in an elastic manner,
absorbing the stress while the outer woven cord simply
expands to a strain of ca. 1.2. At this point the cord is
fully stretched, and it starts to respond as a stiff fiber.
Thus with additional stress there is little strain. This
region has a high modulus. Two regions exist for the com-
posite: (i) a low modulus region early when the rubber
is carrying the load, and (ii) a high modulus region where
the outer stretched cord is carrying the load.
As the data above indicate, well-defined stress/strain
curves are generated using this apparatus. Mention is
made of using “constant”-weight objects, such as bolts,
available at any hardware store. Table 3 shows the pre-
cision in weights of one set of bolts that we used. One
can routinely expect a standard deviation of ca. 0.2%.
Thus in adding weights consecutively as described in the
procedure, we can assume all additions to provide
equivalent weight. This makes these measurements
quite straightforward and fast.
Conclusion
A simple, inexpensive, and quite instructive appara-
tus has been designed to measure tensile properties. It
should be applicable to undergraduate and graduate
courses where polymers, principles of engineering mate-
rials, and conversions of units involving English to met-
ric, force to energy, mass to pressure, etc., are of concern.
Vol. 73 No. 11 November 1996 • Journal of Chemical Education
In the Laboratory
Figure 7. Stress vs. strain for elastic strap.
Acknowledgments
The helpfulness of following individuals is gratefully
acknowledged: Jean Strong at UWSP, Beth McGraw at
RPI, and Janet Lewis at USM for secretarial support and
oversight of preparations for the summer workshops; S.
Czysz, Jason Waldkirch, S. Baird, T. Maeder, J. Opsteen,
Russ Cartwright, Stephanie Fredrickson, Chris Petersen,
C. Roberts, G. Tullos, student assistants at the three
sites; and S. Krause (RPI), Paul Hladky (UWSP), and
Members of the Advisory Panel for their helpfulness.
This project was supported by National Science Foun-
dation Division of Undergraduate Education Grants DUE
#91-50497 and 92-54351 and by POLYED, the joint poly-
mer education committee of the ACS Polymer Divisions.
This experimental procedure was developed as part
of the 1993 NSF-POLYED Scholars Program directed by
John P. Droske of the University of Wisconsin–Stevens
Point. The author was appointed an NSF-POLYED Scholar
for 1993 and gratefully acknowledges the financial sup-
port of NSF in grants DUE #91-50497 and #92-54351.
Most of the confirming data and refinement of the
method are due to the efforts of Matthew Williams, the
primary contributor to this work, and an undergradu-
ate Engineering student at the University of Michigan-
Dearborn. J.P. Droske and one of his students, Stephanie
Fredrickson, assisted in generating some of the confirm-
ing stress–strain data.
These experiments were prepared as part of the
NSF-POLYED Scholars Program. This collaborative ef-
fort involves the following individuals working at three
regional sites:
University of Wisconsin–Stevens Point: John P. Droske, UWSP, Project
Director; Jerry P. Jasinski†, Keene State College, NH; Joe Young‡,
Chicago State University. Rensselaer Polytechnic Institute: Gary
Wnek, RPI Site Director; Karen S. Quaal†‡, Siena College, NY; Chang-
Ning Wu †‡ , University of Massachusetts–Dartmouth. University of
Southern Mississippi: Lon Mathias, USM Site Director; T. Carter
Gilmer†, University of Michigan–Dearborn; Guy Mattson‡, Univer-
sity of Central Florida;E. Ann Nalley†‡ , Cameron University, OK.
†
NSF-POLYED Scholar, 1993; ‡NSF-POLYED Scholar, 1992.
Literature Cited
1. Young, R. J.; Lovell, P. A. Introduction to Polymers; Chapman and Hall: Lon-
don, 1991.
2. Sperling, L. H. Introduction to Physical Polymer Science; Wiley: New York, 1986.
3. Handbook of Plastic Materials and Technology; Rubin, I. I., Ed.; Wiley: New
York, 1990; p 355.
4. Van Krevelen, D. W., Properties of Polymers; Elsevier: New York, 1972.
5. Rodriguez, R. J. Chem. Educ. 1990, 67, 784.
6. Rubin, I. I., Ed. Handbook of Plastic Materials and Technology, Wiley: New
York, 1990.
1065