Engineering Design with Polymers

Society of Plastic Engineers

Ron Rorrer, Ph.D., P.E. Associate Professor University of Colorado Denver [email protected]
Ronald A. L. Rorrer, 2009

Introduction Engineering view of polymers. Fluid Solid Mechanical models of viscoelastic behavior. Fluid Solid Advanced Why Model? Thermomechanical Spectrum Time-Temperature Superposition Boltzman Superposition Viscoelastic correspondence principle FEA Example Creep Rupture Buckling Damping and Isolation Discussion Summary Questions References

Outline

5-6-09

Introduction

Designing with non-polymeric materials

Designing with polymeric materials

Youngs modulus, E, and yield strength, Sy Consider that modulus and strength (yield and ultimate) are now time/rate and temperature dependent Environmental factors can have a much greater impact on polymers than nonpolymers.
5-6-09 3

Introduction (continued)

Engineers must understand the viscoelastic response of polymers in order to use polymers in engineering designs. Mechanical properties will be related to the thermomechanical spectrum. The thermomechanical spectrum is the middle ground and point of commonality between engineers and polymer chemists. Throughout, the thermomechanical spectrum is used as a focal point to discuss a variety of topics of the mechanical response of polymers
5-6-09 4

Engineering View

Polymers are long chain molecules Consider subdividing conceptually into to two groups

Fluid (plate of intertwined spaghetti analogy)thermoplastic fluids Solid (3-D cargo net analogy)-thermosets with covalent bonding between the polymer chains

Of course, somewhat the exception is thermoplastic elastomers, that while not covalently cross-linked act as if the material is cross-linked. However, these pseudo-cross-links only appear this way for short term elasticity.
5-6-09 5

Fluid

Fluid or fluid behavior

Has no long term elasticity

What does this mean?

Under constant loading material will continually deflect. Constant load will eventually fail material. Loads or duty cycle may be such that failure does not occur in application use lifetime. Polycarbonate window panels Polypropylene classroom seating Every student knows one can not sit too long or one will fail!
5-6-09

Caveat

Examples

Solid

Solid behavior

Has long term elasticity

What does this mean?

Caveat

In theory, if loads are not large enough to completely rupture all of the crosslinks, the solid will Asymptotically approach a limiting deflection. Carry a constant load indefinitely. Even though initial loading is carried, crosslinks can rupture in a cascade fashion over time. Every student knows one can not sit too long or one will fail! Epoxy (carbon fiber/epoxy composite) EPDM roofing material Polybutadiene in vehicle tires

Examples

5-6-09

Elements for Modeling

Most elementary models are composed of a collection of springs and dashpots to model the elastic and inelastic aspects of viscoelastic behaivor. = E Spring Dashpot
=

5-6-09

Fluid Model

Simplest Fluid Model

Maxwell Model

Spring in series with damper

Apply constant creep stress to model, monitor strain

0 t

5-6-09

Solid Model

Simplest Solid Model

Kelvin Model

Spring in parallel with damper

Apply constant strain to model, monitor stress

= (t ) 0 E (1 e Et )

Advanced Viscoelastic Models

The Maxwell and Kelvin models are too simplistic for most real world behavior. However, they do capture the essence of fluid and solid behavior. In order to fit the behavior over many decades of time other models are used.

5-6-09

11

N-parameter

Conceptual extension n-parameter models

n-parameter fluid
... ,

n-parameter solid

E (= t ) E0 +

E e
i =1 i

t ti

... ,
5-6-09 12

Spring Pot -Fractional Model

Eo

E1
,

where:

= memory parameter, not necessarily integer E= modulus = viscosity D= fractional differentiation with respect to

E22

=0 and the spring-pot reduces to a simple spring =1 and the spring-pot reduces to a dashpot. 0<<1, the spring-pot contains characteristics of both a spring and a dashpot. Conceptually, this eliminates the limitation of an element being solely either a spring or a dashpot. In fact, this is what leads to the springpot response over multiple time scales, unlike the spring and dashpot models which are constrained by their exponential time dependence.

5-6-09

13

Fit to Experimental Data

12.0

Modulus [MPa]

10.0

8.0

6.0

4.0

2.0

0.0 -5.0

-4.0

-3.0

-2.0

-1.0

0.0

1.0

2.0

3.0

4.0

Log t [log s]

Solid data from Gottenberg and Christiansen1 fit with a series of exponentials and a fractional spring-pot model2
5-6-09 14

Why Model?

Experimental data can be used to calculate response at a single time or frequency However, a model allows analytical or numerical prediction of response over all frequency and times Prior transformed data and models extend from 10,000 Hz to 1000 s.
5-6-09 15

Thermomechanical Spectrum

Shear testing of Nylon 6,6 Notes:

Glass transition temperature of Nylon 6,6 routinely reported as 50 C, here 65-80 C Strain level for Dynamic Mechanical Analysis often below application levels

5-6-09

16

Thermomechanical Regions

Glassy Region

Leathery Region

Material responds rigidly, region of highest storage modulus, low damping Transtion region, highest damping Material has low modulus Thermoplastic fluid melts Crosslinked thermoset solid degrades
5-6-09 17

Rubbery Region

Terminal Region

Importance of Regions

Importance of regions illustrated by ductile to brittle transition, TT, of impact properties. For structural purposes polymers are typically used in either the rubbery or glassy region. A transition between regions during use can be disastrous for performance of a structural material or adhesive.

Impact Strength

Brittle

Ductile

Tg TT Temperature
18

5-6-09

Time-Temperature Superposition (TTS)

Time-temperature equivalence is based upon the observation that temperature and rate are inter-related with respect to polymer mechanical properties. Macroscopically, we can observe that the effect of cold temperatures are similar to the effect of high rates on polymer response.
5-6-09 19

TTS

While TTS is not always valid, there are two shift factors that have been used to take data generated over a 1-3 decade range of rates and create a master curve over 515 decades of rate Applicability (Tg<T<Tg +150 C)
Arrhenius Williams, Landel, and Ferry (WLF)
1 1 b T T 0

aT = e

C1 (T T0 ) log aT = C2 + (T T0 )

Where: T0=reference temperature b, C1, and C2 are constants

TTS Example of Amorphous Polymer3

5-6-09

21

Application of Boltzman Superposition

Results from constant load creep test can be used to predict response to time varying load.
for t t1 fo r t > t1 = ( t ) 0 J ( t ) + 1 J ( t t1 )

= (t ) 0 J (t )

( t )= 0 J ( t ) + J ( t t1 )
0

d i d it dti

5-6-09

22

Viscoelastic Correspondence Principle

The time dependent deformation of a linear situation (both material properties and deformation) can be analyzed by substituting the E(t) or 1/J(t) for E if the static solution is known. Consider a cantilever beam with a transverse end load. ( x) Static deflection y( x) = P f6EI f ( x) f ( x ) J (t ) y ( x, t ) P = P Time dependent deflection = 6 E (t ) I 6I
5-6-09 23

FEA Example

This concept can be extended to FEA. Example

t=0

Since FEA is linear, (0)=0.5(t=t1)

Polyamide 100 mm OD, 30 mm ID Constrained on bottom half of exterior Loaded vertically on inner half of bore with 100 N distributed load Assume that at t1 E(t1)=0.5 E(t=0).

t=t1

5-6-09

24

FEA Example

The maximum stress does not change. This occurs because this problem does not experience non-linear deformation.

t=0

t=t1

5-6-09

25

Creep Rupture

Due to viscoelasticity, polymers will fail under constant load

5-6-09

26

Miners Rule
tn tn 1 1 t n =1 cn
N

Originally used to approximate fatigue failure of metals. In many cases Miners rule has been shown to apply to both creep and fatigue failure of polymers.

5-6-09

27

Creep Buckling

Viscoelastic Correspondence Principle can be extended to buckling Euler Buckling

2 EI
L2 E (1 e Et n )
Maxwell Pcr

Kelvin Solid Pcr =

I
2

I E = 2 L 1 + Et
2

5-6-09

28

Damping and Isolation

Fundamental difference between damping and isolation.

Damping-energy of vibration is dissipated as heat due to viscous losses in material being cycled. Isolation-due to dynamic response of system, when excited at frequencies away from natural frequencies (resonances) of system, lack of system dynamic response isolates driven components
5-6-09

29

Vibrations and the Thermomechanical Spectrum

Damping is one of the few situations where it is desirable to use a polymer near the glass transition. The temperature axis due to time superposition is equivalently
5-6-09

Temperature(f= 1 Hz)

(T = 1 Hz ) (T2 , f 2 ) 1 , f1

E/E= tan

30

Vibration Models

Most Dynamic Mechanical Analyses assume that the material being tested is a Kelvin Solid with the complex modulus being written in the various forms found in the literature as
E * = E '+ iE '' = E '(1 + i ) = E '(1 + i ) = E '(1 + tan )

In the model that follows

k n = m

c 2 km

A= area L=the height of polymer represented by the spring and damper

Damping and Isolation of a SDOF System

k
4

= 0 = 0.1

F0 sin t

XK/F

= 0.25

2 = 0.5 1

= 1.0 0 0 1 2

c k F 2 2 + x = sin t =+ + n x+ x x 2n x x m m m

w/w n

A single dof system has a frequency response function as shown. For w/wn <0.5 and w/wn > 1.4 the system responds as an isolator. Between 0.5 and 1.4 the system is dominated by damping
5-6-09 32

Polyurethane Blends4

Polyurethanes are often used as vibration dampers

Slight changes in processing yield large changes in properties Blends of polyurethane can shift peak damping relative to frequency
5-6-09 33

Anamolous Blend

Most blends, have a Tg between the two materials or evidences two Tgs Of interest is a blend of polyisoprene (PIP) and polybutadiene (PBD) which has a peak in losses outside of this range.5
5-6-09 34

Discussion

Polymers have all of the same considerations for design as nonpolymers. The time dependence of viscoelasticity requires additional consideration in the design process.

5-6-09

35

Summary

Design engineers must be aware of the differences between polymers and nonpolymers to create successful designs.

5-6-09

36

Questions?

5-6-09

37

Acknowledgement

The material for this webnair largely abstracted from Gerdeen, J., Lord, H., Rorrer, R. A.L. Engineering Design with Polymers and Composites, CRC Press, Boca Raton, Florida, 2005, 349 pages.6

5-6-09

38

References
1 2

Gottenberg, W. G., and Christiansen, R. M., Int. J. Eng. Sci., vol. 2, pp. 4556, 1964 Welch, S. W. J, Rorrer, R. A. L., and Duren Jr., R. G., Mechanics of TimeDependent Materials, Vol. 3, No. 3, pp. 277-301, 1999 Ferry, J.D., Viscoelastic Properties of Polymers, John Wiley & Sons, New York, 1980. Corsaro, R.D. and Sperling, L. H., Eds., Sound and Vibration Damping with Polymers, ACS Symposium Series 424, J.V. Duffy et al., p. 292, 1990. Corsaro, R.D. and Sperling, L. H., Eds., Sound and Vibration Damping with Polymers, ACS Symposium Series 424, C. M. Roland and C.A. Trask, p. 310, 1990. Gerdeen, J., Lord, H., Rorrer, R. A.L. Engineering Design with Polymers and Composites, CRC Press, Boca Raton, Florida, 2006, 349 pages.

5-6-09

39