Ind. Eng. Chem. Prod. Res. Dev. l984, 23,426-434
426
GENERAL ARTICLES Material Characterization of Structural Adhesives in the Lap Shear Mode. 1. The Effects of Rate Erol Sancaktar,’ Steven C. Schenck, and Shrkan! Padgllwar Department of Mechanical and Industrlai Engineering. Clarkson College of Technology, Potsdam, New York 13676
A general method for characterizing structural adhesives in the bonded lap shear mode is proposed. This first paper evaluates the proposed characterization method wlth the use of experimental constant strain rate data. Four different model adhesives are used in the bonded lap shear mode for this purpose. The model adhesives are FM 73 and FM 300 modified epoxy adhesives and LARC 3 and thermoplastic polyimidesuifone linear high molecular weight polyimide adhesives. Elevated temperature behavior is also studied. Experimental results reveal that it is possible to describe the constant straln rate shear stress-strain behavior of structural adhesives in the bonded form by using viscoelastic or nonlinear elastic relations. Ludwik’s equation provides an adequate description of the rate dependent ultimate and elastic limlt shear stresses. I t is also concluded that the Thermoplastic Polyimidesulfone adhesive is superior to the epoxy adhesives studied In high temperature strength retention. Results on the creep and delayed failure behavior of the model adhesives will be presented in a subsequent paper.
Introduction Structural adhesives are preferred over the customary penetration techniques for joining of modern lightweight-composite materials. Their usage increased considerably during the past decade in the aerospace, automotive, naval, and many other industries. Structural adhesives are superior to conventional means of joining materials because of their flexibility and toughness, lightness of weight, exceptional thermal stability, solvent and moisture resistance, and excellent mechanical properties at room and elevated temperatures. In joining mechanical components, the structural adhesive and mode of bonding (i-e., lap, butt, scarf, etc.) to be used is usually determined by in-service requirements. Single lap joints are widely used due to their applicability in many industrial designs. In the absence of catastrophic crack propagations, the failure of lap joints may occur in one of the following modes: (1)rupture of the adhesive layer, when the ultimate stress is reached; (2) creep rupture of the adhesive layer when a high level of constant load is used; (3) failure of the adherends. Failure in the third mode will be unlikely when metal adherends such as steel or titanium are used. It is necessary, however, to consider the first two modes of failure for the design of adhesively bonded joints. Adhesive rupture in the constant strain rate mode will be discussed in this paper. Discussion of the delayed failure phenomenon will be presented in a subsequent paper. It should be noted that the failure of lap joints may also occur by interfacial failures if weak interphases are present. It is usually assumed by researchers such as Jennings (1971) that when proper surface preparation is provided the adhesiveladherend interface becomes stronger than the bulk of the adhesive. If interfacial impurities are introduced prior or subsequent to the cure process weakinterfacial failures are likely to result. The impurities
o I96-432I1841I223-042t3$0I 5010
which appear after cure processes are the results of chemical environments causing catalysis at the interface. The most common impurity induced in that manner is corrosion. Impurities may also be present on the adherend or adhesive surfaces (this is more likely, as the adherend surfaces are almost always cleaned) prior to bonding. During a previous investigation by Dwight et al. (1980) EDAX (energy dispersive analysis of X-ray fluorescence) examination of the fracture surfaces of aluminum-Metlbond 1113 specimens indicated the presence of silicon on some of them. Silicon caused weak-interfacial failures in these specimens. Hata (1971) suggests that interfacial forces are essentially elastic. Furthermore, since it is theoretically possible to design joints for which the interfacial region is not the weakest, further discussion of interfacial failures will not be presented in this paper. For design purposes, one often needs only the elastic properties (namely Young’s modulus and elastic limit stress) and failure stresses as a function of rate and temperature. If such is the case, then a satisfactory characterization can be obtained with the use of a semiempirical approach to describe the failure stresses as a function of rate and temperature for constant strain rate loading. This approach assumes that the viscoelastic effeds are negligible in the initial portion of the stress-strain curve, so that an initial elastic strain rate can be defined. Previous work by Brinson (1974,1975,1980) Chase and Goldsmith (1974), and Sancaktar (1980, 1982) on a variety of polymeric materials and adhesives proved this assumption to be a valid one. If information on the magnitude of strains (past the elastic limit) is also required, then one needs to use a theoretical approach with the application of a mechanical model to characterize the material. Information on failure stresses can also be extracted from such an approach if 0 1984 American Chemical Society
Ind. Eng. Chem. Prod. Res. Dev., Vol. 23,No. 3, 1984 427
perfectly plastic flow is present (i.e., if the constant strain rate stress-strain curve has an asymptote). It should also be noted that for some linear thermoplastic adhesives the effects of rate on the failure stresses may be negligible enough to permit the use of a nonlinear elastic relation to characterize the constant strain rate stress-strain behavior. The semiempiricalrelation proposed to describe rupture stresses at room and elevated temperatures is Ludwik’s equation for the constant strain rate condition. It should be noted that the original form of Ludwik’s equation as reported by Thorkildsen (1964) does not include thermal effects. This paper proposes an empirical modification of Lukwik’s equation to describe the effects of temperature. Temperature has a strong influence on the mechanical stress-strain properties of an adhesive. For instance, the effects of temperature on the strength of lap shear specimens bonded with a high molecular weight thermoplastic adhesive was investigated by Bugel et al. (1961). They reported a sharp decrease in the joint strength at temperatures of about 71 “C. Experimental data which will be presented in this paper were obtained with the use of the following model adhesives. (1)FM 73 is a modified epoxy adhesive film with polyester knit fabric carrier cloth. It is manufactured by the American Cyanamid Company. Its product information brochure reports a service temperature range of -55 to 120 “C with high moisture resistance. The same adhesive is also available with random polyester mat carrier under the FM 73M brand name (to offer better handling characteristics). (2) FM 300 is a modified epoxy adhesive manufactured by the American Cyanamid Company. It is supported with a “tricot-knit” polyester carrier. The manufacturer’s product information brochure reports superior metal-tometal peel strength and service temperatures to 150 “C along with moisture and corrosion resistance. It is opaque to X-ray. The same adhesive is also available with “random mat” polyester carrier under the brand name FM 300M and with “wide-open knit” polyester carrier under the brand name FM 300K. The use of “random mat” carrier reduces the tendency to trap air during cure and provides good bondline and flow control. The manufacturer’s brochure also reports that the highest overall performance is obtained with the use of “wide-open knit” carrier. (3) LARC 3 is a linear high molecular weight polyimide adhesive developed at NASA Langley Research Center by St. Clair and Progar (1975) for high temperature applications. It is prepared in tape form with style 112 E glass cloth (A-1100 finish) and aluminum powder filler. It can be used in thermal environments with temperatures up to 288 “C. (4) Thermoplastic Polyimidesulfone is a novel adhesive developed at NASA Langley Research Center by St. Clair and Yamaki (1982). It has thermoplastic properties and solvent resistance. It is prepared in tape form on 112 E-glass which has an amino-silane surface treatment. It can be used in thermal environments with temperatures up to 232 “C. Adhesive manufacturers generally do not supply detailed information on the rate, time, and temperature-dependent variation of the lap shear properties. When data are available, they are usually limited to the ultimate strength values. Experimentally measured lap joint strength values for LARC 3, Thermoplastic Polyimidesulfone, FM 73, and FM
300 adhesives are available in the literature as reported by their developers. Information in regard to the rate and temperature-dependent single-lap mechanical behavior for these adhesives will be presented in this paper. Some mechanical data on the Thermoplastic Polyimidesulfone adhesive are available in a paper by St. Clair and Yamaki (1982). They report that the lap shear strength decreased from 28.6 MPa to 18.1 MPa when the test temperature was increased from the room condition to 232 “C. St. Clair and Yamaki also report that thermal aging of the adhesive in the bonded form at temperatures up to 232 “C resulted in a reduced lap shear strength. For example, the ambient temperature strength of lap shear specimens that had been aged for 50 000 h at 232 “C was 25.1 MPa, while the same strength for the unaged specimens was 28.6 MPa. The present investigation uses single-lap specimens which were supplied by the NASA Langley Research Center. For such specimens the shear stresses are calculated as load divided by the overlap area, as prescribed by the ASTM standards. It should be noted, however, that the shear stresses calculated in this manner provide only approximate values as the actual shear stress distribution along the overlap area is not uniform but in fact is part of a biaxial stress state. For this reason, the failure stresses reported in this paper should only be used in comparison to other failure stress values obtained using the ASTM standards. The exact form of the stress distribution in single-lap joints has been described by Goland and Reissner (1944). Analytical Considerations The rate dependency of the rupture stresses under constant strain rate loading can be expressed with the semiempirical equation proposed for metals by Ludwik in the form 7,lt
=
7’
+ 7” log
(;)
where 7dt is the ultimate shear stress, ? is the initial elastic shear strain rate and r’, f ’, and 9’ are material constants. This equation was used successfully by Brinson et al. (1974, 1975,1980) in describing the rate dependence of rupture stresses for polymeric materials in the bulk tensile and shear modes. The same form of eq 1can also be used to describe the variation of elastic limit shear stress (e) and strains (4) with initial elastic shear strain rates. These expressions may be written as B = 0’ 8” log ( + / y ) (2)
+
and (3) 4 = 4’+ @’log (?/-j,’) where additional materials constants are defined accordingly. Superposition of temperature effects on eq 1,2, and 3 would require experimental justification. For example, examination of rate vs. strength data for copper at various temperatures (23 OC < T < lo00 “C)as reported by Nadai and Manjoine (1941) indicates a near parallel shift in the strain rate vs. strength relations particularly in the 200 to 800 “C region. Such a parallel shift in eq 1,2, and 3 could be expressed as 7ult = {aT)[7’1 + 7”log (?/?’I (4) 0 = (bT)[0’]+ 0’’ log (+/-y) (5) (6) 4 = k T j [ 4 l + 4’%? (‘?/?’I
428
Ind. Eng. Chem. Prod. Res. Dev., Vol. 23,No. 3, 1984
4 = (+/Go) + (7 - T s ) / Q (7 > 7 8 ) = Go@+ Gly)/(Go + G,) and tj = Gos/(Go+ G,).
i l l '
where T, To obtain the constant strain rate relation, the above constitutive equations are solved according to the condition given by eq 8 to result in
I
Y Chase-Soldsmith
Model
7
= Goy
(7
h0k9+ vR + G ~ r ) ( -l exp[-a(y
IT
~ )
- 4)l)}/(Go + GI) + [8 - 611 (7 > T J
+ GoGl(r - b ) / ( G o+ GJ1 exp[-4y
(11)
where a = G0/qR. If the viscoelastic effects on adhesive behavior are determined to be weak, then a nonlinear elastic relation can be used to fit the stress-strain data. The power function relation expressed for the state of pure shear in the form
lp.
n Madified Bingham Model
7
V
J y,jTPAIN
Figure 1. Viscoelastic models used to describe adhesive behavior.
where UT, bT, and CT are functions of temperature. The present investigations on three different thermosetting and thermoplastic adhesives with carrier cloths revealed that viscoelastic modified Bingham and ChaseGoldsmith models could be used successfully to characterize the constant strain rate stress-strain behavior of the adhesives in the bonded form. The modified Bingham model was first used by Brinson et al. (1974, 1975) for characterizing the bulk tensile behavior of polycarbonate and Metlbond adhesives. The same model was later used by Sancaktar and Brinson (1980) for characterizing the bulk shear behavior of Metlbond adhesives. In order to be valid for the pure shear mode, the one-dimensional constitutive equation for the modified Bingham model was applied in the form
4= i/G 4 = ( F I G )+ (7 - e ) / r
(7
(7)
5 8)
(e < I 7 u 1 ~
where T and y are shear stress and strain, respectively. A sketch of the mechanical model is shown in Figure 1. The constant strain rate stress-strain relation based on the modified Bingham model is obtained by solving the constitutive eq 7 according to the condition
4 = R = constant
(8)
to result in T = ~ Y 7
=e
+P
( 7 ~ e )
- exp[-(y - d G / ~ Y l l
(9) (0
0-1
-
26 9
+0
liirec
25
9 1 l o p (-)
I
I
102
101
100
30
0
I\ITIA1 ELASTIC F T R A I N RATE
Isec)
Figure 17. Variation of ultimate shear stress with initial elastic strain rate and comparison with Ludwik's equation for Thermoplastic Polyimidesulfone adhesive.
80
c-
i(%= ) Y* '
38.97
+
0 0
EXPERIMENTAL DATA
5.64 l o g (x.j
D A T A YO? INCLLTED IS CL!VE FITTISG
= l%/sec
0
0
0
- o0 I 10'1
100
I
I
:01
IO2
I N I T I A L E L A S T I C STRAIN R 4 T E (%/set)
Figure 18. Variation of maximum shear strain with initial elastic strain rate for Thermoplastic Polyimidesulfone adhesive.
adhesive-matrix to adhesive-fiber and/or adhesiveadherend interfacial) of the lap joint. Conclusions The present investigation was concerned with the development of a practical method for characterizing structural adhesives in the bonded lap shear mode. The validity of this proposed characterization method was evaluated with the use of experimental constant strain rate, creep, and creep-rupture data. This paper presented experi-
434
Ind. Eng. Chem. Prod. Res. Dev., Vol.
23, No. 3, 1984
1
"
1'-
perature strength retention. Results on the creep and delayed failure behavior of the model adhesives will be presented in a subsequent paper. Acknowledgment This work was supported by NASA Langley Research Center under NASA Grant NAG-1-284. Dr William S. Johnson was the NASA technical monitor and his assistance in the conduct of this work is gratefully acknowledged. The authors also wish to express appreciation to Dr. Terry L. St. Clair of NASA Langley for his guidance and assistance throughout the implementation of this project. Registry No. FM73,60181-90-0; FM300,71210-48-5; LARC 3, 67339-99-5; (BTDA).(3,3'-DDS) (copolymer), 28825-50-5; (BTDA)*(3,3'-DDS) (SRU), 54571-77-6.
Literature Cited
1
5
On
20
"0
60
n
SHEAR STRAIY (Z)
F i g u r e 19. T h e effects o f temperature on t h e constant s t r a i n r a t e stress-strain behavior of Thermoplastic Polyimidesulfone adhesive.
mental results on the constant strain rate behavior. Four different model adhesives were used in the bonded lap shear mode for this purpose. Elevated temperature behavior was also studied. Based on the experimental data from the four adhesives studied, the following conclusions can be made. (1) It is pasible to describe the constant strain rate shear stress-strain behavior of structural adhesives in the bonded form by using viscoelastic or nonlinear elastic relations. (2) Ludwik's equation provides an adequate description of the rate dependent ultimate and elastic limit shear stresses. Such behavior indicates that when a joint is subjected to loads at varying rates, the lowest rate-strength should be used in the strength design criterion. (3) The epoxy adhesives are inferior to the polyimide adhesive, Thermoplastic Polyimidesulfone, in high tem-
American Cyanamid Co. FM 73 Adhesive Film Product Information Brochure; Havre de Grace, MD 21078. American Cyanamid Co. FM 300 Adhesive Film Product Information Brochure; Havre de Grace, MD 21078. Brinson, H. F.; Exp. Mech. 1970, 70(2),72. Brinson, H. F.; Renieri, M. P.; Herakovich, C. T. I n "Fracture Mechanics of Composites", ASTM STP 593, 1975; p 177. Bugel, T. E.; Norwalk, S.; Snedeker, R. H. I n "Phenoxy Resin-A New Thermoplastic Adhesive", Eley, D. D., Ed.; Oxford University Press, 1961; p 87. Chase, K. W.;Goldsmith, W.;Exp. Mech:1974, 74(1), 10. Dwight, D. W.;Sancaktar, E.; Brinson, H. F., I n "Adhesion and Adsorption of Polymers", Lee, L. H., Ed.; Plenum: New York, 1980; Vol. 12-A, p 141. Goland, M.; Relssner, E.; J. Appl. Mech. 1944, 7 1 ( 1 ) , A-17. Hata, T. I n "Recent Advances in Adhesion", Lee, L. H., Ed.; Gordon and Breach Science Publishers, 1973; p 269. Jennings, L. W. I n "Recent Advances in Adhesion", Lee, L. H., Ed.; Gordon and Breach Science Publishers, 1973; p 469. Mall, S.; Johnson, W. S.; Everett, R. A., Jr. "Cyclic Debonding of Adhesively Bonded Composites", NASA Langley Research Center, Hampton VA, 1982; NASA Report No. 84577, p 6. Nadai, A.; Manjoine, M. J.; J. Appl. Mech. 1941, 8, A-82. Sancaktar, E.; Brinson, H. F. I n "Adhesion and Adsorption of Polymers", Lee, L. H. Ed.; Plenum: New York, 1980; Vol. 12-A, p 279. Sancaktar, E.; Padgilwar, S.; J. Mech. Des. 1982, 704(3), 643. St. Clair, T. L.; Progar, D. Pokm. Prepr. 1975, 16(1), 538. St. Clair, T. L.; Yamaki, D. A. "A Thermoplastic Polyimidesulfone", NASA Langley Research Center, Hampton, VA, 1982; NASA Report No. 84574. ThorkiMsen, R. L. I n "Englneering Design for Plastics", Baer, E., Ed.; Reinhoit: New York, 1964; p 295.
Receioed for review N o v e m b e r 22, 1983 Revised manuscript received February 24, 1984 Accepted M a r c h 12, 1984
