METHOD
OF
TESTING
GLASS
G. S. Pisarenko, Yu, and G. M, Okhrimenko
IN M.
MONOAXIAL
COMPRESSION
Rodichev,
UDC
539.411:620.173 .24:666
Glass, possessing high compressive strength, is a promising material for components working under compressive load conditions. Considerable difficulties in using glass as a structural material are related to the absence of reliable data on the actual strengths of commercial glasses accepted by industry. According to reference data the compressive strength of such glasses is 50-200 kgf/mm 2, whilst the deviation from the mean values is ~50% [1]. For use in highly stressed structures a glass is required which has a strength of 200 kgf/mm 2 and above. The absence of reliable and c o r r e c t methods of determining the compressive strength gives m i s leading conceptions of the structural properties of glass, which makes it difficult to select the type for the production of critical components. Methods are also unknown for obtaining stable high glass strengths during compressive testing. It is thought that high strengths can only be attained under special laboratory conditions, for example under vacuum, with careful preservation of the testpiece surfaces etc. [2]. Several methods are proposed in the present paper for improving the compressive testing of c o m m e r cial glasses, which provide a means of obtaining strengths of up to 250 kgf/mm 2 for types accepted by industry, this being several times g r e a t e r than the data in the reference literature. The compressive strength of glass is usually determined on prismatic or cylindrical testpieces, loaded by means of fiat steel bearings. The State Institute of Glass has recommendedcubic o r p r i s m a t i c testpieces of dimensions 4 • 4 • 4 and 4 x 4 • 6 mm [3]. In the Institute of Strength of Materials , Academy of Sciencesof the Ukrainian SSRandthe "Avtosteklo" Scientific Research Institute cylindrical testpieces of diameter up to 10 mm and prismatic testpieces with a t r a n s v e r s e section of up to 10 x 10 mm and height 20-40 mm were used. These provided a means of showing the structural properties of the glass more accurately and in
Is
Is
a
b
Fig. 2 Fig. 1. Steel bearing with t r a c e s of the embedded glass 13 V for a compressive s t r e s s equal to approximately 116 kgf/mm 2. Temper 3/4. Fig. 2. Diagram of composite: a) cylindrical and tubular and; b) p r i s matic glass testpieces for monoaxial compression testing. Institute of Strength of Materials, Academy of Science s of the Ukrainian SSR, Kiev. Translated f r o m Problemy Prochnosti, No. 10, pp. 17-22, October, 1973. Original article submitted January 15, 1973. 9 1974 Consultants Bureau, a division of Plenum Publishing Corporation, 227 West 17th Street, New York, N. Y. 100II. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, microfilmi~g, recording or otherwise, without written permission of the publisher. A copy of this article is available from the publisher for $15.00.
1177
3
7L_ P Fig. 3
Fig. 4
Fig. 3. Diagram showing normal method of loading for testpieces of constant cross section: I) glass rod; 2) and 3) smooth and spherical bearings of type ShKh-15 steel; 4) spherical copper thrust bearing. Fig. 4. Attachment for grinding the ends of the glass rods: i) constant cross section glass rod; 2, 3) half rings; 4) pacldng; 5) faceplate ShPZ35OIVi machine (5 is the single ended allowance in treating the end; co is the rate of rotation of the machine faeeplate). particular, applicable to large components,
to raise the accuracy of test-piece loading.
During the testing of such testpieces on flat bearings the failure is most frequently initiated from the end surfaces even at a loading of 50% of the limiting value. This is due to the exceptionally unfavorable complex stress -- strain distribution of the material at the ends of the testpiece, characterized by considerable non-uniformity, which is dangerous for brittle materials very sensitive to stress concentration [4, 5]. The actual pattern of stress and strain in the bearing zone during loading by fiat bearings is a ftulction of the accuracy of testpiece and bearing preparation, the strength of the test material, the hardness, strength and rigidity of the bearings, the rigidity of the loading equipment, and also the relation between the Poisson ratio and Young's modulus, determining the transverse strain of the ~estpiece and bearings [6]. The indefinite and unfavorable natures of the bearing condition are made worse by the fact that, for mean calculated compressive stresses of 60-80 kgf/mm 2 stress concentration causes the high strength hardened steel supports to be strained into the plastic range round the perimeter of the bearing section, At higher stresses there is a mutual embedding of the contacting layers of testpiece and bearing (Fig. i), this being related to the danger of cleavage of the surface flaws which are always presented in testpieces of commercial glass. The damage occurring in the bearing surface zone during the course of loading imposes an imprint on the stress -- strain distribution of the testpieces, their resistance to strain and to failure, and consequently on the final test results. Most frequently reduced results are obtained which do not reflect the actual strength of the glass in the working part of the testpieee, and being characterized by considerable scatter, which in some cases is not considered as a characteristic of the test method but of the properties of the material investigated. Evidently, in order to determine the actual strength of the glass under compression it is necessary to ensure that the failure process is generated first of all in the working zone free from boundary effects, and to achieve monoaxial compression in this zone. For this purpose it is necessary to use such methods of loading the testpiece which considerably raise the bearing capacity of the support zones. It is possible to resolve this problem either as a result of increasing the area of the tes0piece bearing surface and loading them more uniformly, or by developing special measures to reduce the stresses of the end surfaces and operate the bearing zones in the triaxial compression zone.
1178
TABLE 1 Test method
On plane bearings
In rir~gs
TestExtemaldiapieeeNo, meter D, mm
Cross section area F, mm2
Compressive stress at time of failure, k~f/mm 2
Rods of type 13v lass * 50,869 98 64,150 lll 117 64,435 ll0 65,291 116 S0,616 110 50,500 89 58,058 52,525 95
Remarks
The cylindrical surface of the testpiece was mechanicallydamaged by means of medium grade abrasive
2 3 4 5 6 7 8
8,05 9,04 9,06 9,12 8,03 8,01 8,60 8,18
9 l0 11 12 13
9,38 9,36 9,40 8,25 8,66
69,067 68,773 69,362 53,429 58,871
237
14 15
9,25 9,60
67,166 72,345
189 163
Type of damage similar to testpieces Nos. 1-8
1
232 245
The cylindrical surface was flame polished
237
253
Qua~z glass tubes On plane bearings
16 17 18 19 20 21
10,06 9,94 10,20 10,13 10,10 10,15
65,531 65,581 69,70 67,422 66,800 66,694
115 106 110 115 I10 116
The cylindrical surface was flame polished
In tings
22 23 24
I0,I1 10,07 10,14
66,717 65,952 66,534
134 125 I38
Rings glued with sealing wax
25 26 27 28
10,12 10,05 1,0,08 9,91
66,482 65,041 65,848 63,639
II2 124 144 156
Epoxy resin and testpiece not heat treated. Testpiece surface fLame polished
29 30 31 32 33
10,12 10,15 10,16 10,05 i0,08
66,349 67,022 66,786 65,175 65.582
201 210 170 258 211
Epoxy resin and testpiece heat treated. Testpiece surface heat treated
*According to the data in[3] the linear compressive strength of type 13 v glass is equal to 66.6 kgf/mm2.
The f i r s t m e t h o d , u s e d s u c c e s s f u l l y i n [7], i s c o u p l e d with the p r e p a r a t i o n of t e s t p i e c e s of v a r i a b l e c r o s s s e c t i o n , h a v i n g w o r k i n g z o n e s of r e d u c e d t r a n s v e r s e d i m e n s i o n s i n c o m p a r i s o n with the b e a r i n g s e c t i o n s . The u s e of such test-pieces u n d e r i n d u s t r i a l c o n d i t i o n s i s e x t r e m e l y l i m i t e d due to the c o n s i d e r a b l e technical difficulties. F o r t e c h n i c a l c o n s i d e r a t i o n s p o l i s h e d c y l i n d r i c a l , t u b u l a r , a n d p r i s m a t i c t e s t p i e c e s w e r e u s e d i n the p r e s e n t work. In o r d e r to r a i s e the b e a r i n g c a p a c i t i e s of the t e s t p i e c e e n d s they w e r e f a s t e n e d i n s u i t a b l e h o u s i n g s of the b e a r i n g s b y m e a n s of m a t e r i a l s h a r d e n i n g a t n o r m a l t e m p e r a t u r e ( s e a l i n g wax a n d cold s e t ting epoxy resins). F i g u r e 2 shows d e s i g n v a r i a n t s of t e s t p i e c e s m a d e f r o m c y l i n d r i c a l , p r i s m a t i c , a n d t u b u l a r g l a s s r o d s . A v i r t u e of the p r o p o s e d m e t h o d i s t h a t the h a r d e n i n g l a y e r of wax o r a d h e s i v e i n the c l e a r a n c e s b e t w e e n the b e a r i n g s u r f a c e s p r o v i d e s a s u b s t a n t i a l e q u a l i z a t i o n of the c o n t a c t s t r e s s e s . The t h i c k n e s s of the a d h e s i v e f i l m w a s s e l e c t e d a s the m i n i m u m (0.1-0.15 ram) f o r the c o n d i t i o n that t h e r e i s no d i r e c t c o n t a c t b e t w e e n g l a s s a n d m e t a l , a n d a l s o that the f i l m i s n o t s q u e e z e d out a t m e a n c o n t a c t p r e s s u r e s of up to 100-150 k g f / m m 2. The h o u s i n g w a s m a d e such that the side c l e a r a n c e s , a l s o f i l l e d with a d h e s i v e , w e r e a p p r o x i m a t e l y 0 . 1 5 - 0 . 2 m m on the side. By i n c r e a s i n g the depth of the h o u s i n g , a n d c o n s e q u e n t l y a l s o the a r e a of the side a d h e s i v e l a y e r , the s t r e s s l e v e l of the b e a r i n g s u r f a c e s m a y be
1179
TABLE 2 Test method
On plane bea~ngs
In rings
bearings Fig. 2).
Compressive 2 ITest- Side of Cross section[ stress, kgf/mm [piece base a, I at ttm time ot area F, fai~[- ~ No. mm mm2 1~** anee of Iure first crack 9,94 9,98 0,07 9,96 9,98 9,98 0,07 0,05
98,7O42 99,0016 100,8007 98,6040 99,4008 98,5026 101,3042 100,9200
61 41 71 78 69 56 65 53
9,98 0,05 0,07 0,03 0,08 0,06
99,5000 1~0,0980 100,1960 100,0990 101,9030 I00,4990
136 107 124 109 80 99
considerably reduced. The side adhesive layer also prevents the adhesive being pressed out from the clearances between the bearing surfaces, thereby creating a spatial stress distribution. Since the contact pressure s with te stpiece strengths of more than 100 kgf/rnm 2 are several times greater than the yield point of the adhesive the bearing layer functions as a hydrostatic cushion which provides uniform testpiece loading. An advantage of the side adhesive layer is that at large stresses it limits to a certain extent the transv e r s e strain of the testpieee ends, creating in them a triaxial compression which prevents the possibility of tensile stresses arising in the ends.
In order to obtain experimental verification of the proposed solution polished cylindrical, prismatic, and tubular testpieces of type 13v glass and electro-vacuum type $49-2, and also tubing of transparent quartz glass (GOST 8680-58) were tested by a normal method on flat (Fig. 3), and on bearings attached to the testpieces by wax and cold curing epoxy adhesive (see
The height h of the glass rod for the tested specimens on plane bearings and in the rings was respectively for the type 13v glass 2D and 3D, for the quartz glass tube it was 2.513 and 3.5D, and for the type $49-2 glass it was 2a (see Figs. 2 and 3). The circular and tubular testpieees had flame polished side surfaces, no special measures being taken to preserve these. These testpteees were cut from blanks by synthetic diamond wheels on an M1 binder with copious water cooling of the cutting zone. The attachment for abrasion treatment of the ends on the grinding and polishing machine, used to provide accurate geometry V8 surface finish, is shown in Fig. 4. The rectangular rods of type $49-2 glass were machinedover all the flat faces, in the department of glass and mierocrystalline glass technologyofthe Gor'kiiPolytechnic Institute, to a class V8 finish, and were then etched in a mixture of hydrofluoric and sulfuric acids. The sharp edges of the testpieces were not chamfered. Before assembly, all the components and the testpieces were carefully degreased with alcohol. First of all the flat bearing 4 and the ring 5 were assembled, then a film of epoxy resin was put on the socl~et wall and the end of the glass rod and the rod was fitted into the socket. In a similar way the upper ring 5 and the spherical bearing 3 were assembled. The prismatic rod 1 (see Fig. 2b) was glued in a similar sequence, into the closed end bearings 2 and 3. Plane parallel length gage units were inserted between the surfaces A and B (see Fig. 2a). The adhesive hardened over a period of 4-5 h at room temperature and with a load of i. 5 kgfapplied to the components 3 and 4. The adhesive joint was later heat treated at a temperature of 130 • 5~ for 14-16h In order to increase the strength of the adhesive. Some deviations from the stated methods of preparing the testpieces are shown in Tables 1-3. The testpieees were loaded on equipment within the range 0.35- 0.4 kgf/mm 2.
described earlier in [8], the loading rate being maintained
Due to the good fitting and careful load centering the testpieces tested with plane bearings had on average almost double the strength quoted in reference books. The scatter for the strengths of the testpieces with flame polished surfaces was within the range from --7.5 to +10%. For prismatic testpieees of type $49-2 glass an even greater scatter was noted (from --31 to +21%), this evidently being related to mechanical damage of their side surfaces. In this case they had all the disadvantages stated above. It should be noted that the cylindrical and tubular testpieces with flame polished surfaces usually had less damage during loading, whilst the visible splinters appeared in the stress ranges close to the limiting values. In the prismatic testpieces of type S49-2 glass with the damaged surface the cracks were produced very early (see Table 2) and included a much greater zone adjoining the bearing surfaces. Sometimes
1180
TABLE 3 Test method
On plane
bearings
s 49-'i""~ia~s Glass tubes . ..... 13v ~ s s strength, deviation, strength,[ deviation, [strength, deviation,% kgf/mm ~ % kgf/mm21% /kgf/mra= (11o%)
--7,5-
(34794i) --31,6-- +21,5 Srde surfacesmechanically damaged
+10
(1~2~) - 5 , 5 - +Z7 In rings
Remarks
241 --3,7- -+-5 210 -19 -+222 (9--13) (29--33) I 1_7.61 --+7,26 (413747) ~18--+15,4 (1 5)
Cylindrical surfaces flame polished The same Side surfaces mechanically damaged
132
--5,3 - - +4,5
Sealing wax used as adhesive. Side surfaceflame polished
I34 (25--28)
• 16,4
Testpiece glued with epoxy resin without heattreatment.
(22--24)
Cylindrical surface flame polished Note..__.The figureswithin brackets denote the testpiecenumbers used to obtain the mean strength.
b e f o r e failure longitudinal c r a c k s p a r t e d the testpieee into two o r s e v e r a l p a r t s , but it continued to ~4thstand i n c r e a s i n g loads. The inadequate strength of t e s t p i e c e s under such loading is indicated by the sizes of the glass f r a g ments a f t e r failure, the l a r g e s t of which are 0.5-2.5 ram. The compound cylindrical and tubular test]pieces r e s i s t e d strain and failure in a different manner. Up to the limiting load there a r e no visible splinters, and c r a c k s caused by m o r e minute damage were not noted. The failure was of an explosive n a t u r e , without any warning effects. The test-pieces were c r u s h e d t o e x t r e m e ly fine powders, f r e q u e n t l y being found in the a i r in a suspended state. In all the t e s t p i e c e s tested in this way only the working p a r t of the rod was crushed. p a r t s of the testpieces, fastened in the housings, r e m a i n e d whole even a f t e r failure.
The b e a r i n g
The strength of the composite p r i s m a t i c t e s t p i e e e s tested by the method d e s c r i b e d was on average twice that of ~ s t p i e c e s t e s t e d on plane bearings. Evidently, a s also during bending, the p r e s e n c e of abrasion damage on the side s u r f a c e s of the t e s t p i e c e s r e d u c e s the strength of the glass in compression. This c o n f i r m s the test r e s u l t s f r o m the t e s t p i e c e s (Nos. 14 and 15) of type 13v glass, the flame polished s u r f a c e s of ~:hich were abraded by m e a n s of s}~thetic corundum polishing cloth. The strength of such t e s t p i e c e s was reduced by a p p r o x i m a t e l y 35% (see Table 1). It was c h a r a c t e r i s t i c that at s t r e s s e s of 90-95% of those causing failure a c r a c k was o b s e r v e d , the intensity of which i n c r e a s e d as failure approached. Values are given in Table 3 for the mean strength during c o m p r e s s i o n of the glasses under test, these values indicating that the actual strengths of the c o m m e r c i a l g l a s ses are 2-2.5 t i m e s g r e a t e r than the data given in the r e f e r e n c e l i t e r a t u r e . Accordingly, the mean strength of the type 13v glass was 240 k g f / m m 2, and for quartz glass tubing 210 kgf/mm 2. It was also significant that, for a strength of up to 250 k g f / m m 2 for type 13v glass, the s c a t t e r f o r the values d e c r e a s e d to --3.7 to +5%. The method d e s c r i b e d also provided a m e a n s of reducing the s c a t t e r f o r p r i s m a t i c te stpieces. F r o m the r e s u l t s of testing t e s t p i e c e s of different c r o s s section shapes one m a y a s s u m e that the c y lindrical t e s t p i e c e s are to be p r e f e r r e d , since their s t r e s s distribution is m o r e uniform in c o m p a r i s o n to the p r i s m a t i c and tubular testpieces. F u r t h e r m o r e , the cylindrical t e s t p i e c e s are m o r e suitable since the glass r o d s used in t h e i r p r e p a r a t i o n a r e e a s i l y drawn f r o m the m a s s of glass flowing towards the w o r k piece production. To a significant degree the p r o p o s e d method of testing glass is dependent on the c o r r e c t selection of adhesive m a t e r i a l f o r the ~ g . A low strength adhesive composition may not provide sufficient load b e a r ing capacity for the ends, leading to low results. This is indicated by the test r e s u l t s f o r the tubular 1181
testpieces (see Table I), bonded in the rings by means of sealing wax and epoxy resin without heat treatment (testpieces Nos. 22-28). The results obtained may also be used for the development of highly stressed joints in constructional elements of glass and metals, in which other types of joint do not provide the load bearing capacity required of the component. CONCLUSIONS I. Testing based on plane bearings does not give a full representation of the strength of commercial glass under monoaxial compression. 2. The proposed method of testing compound glass testpieces, based on increasing the load bearing capacity of the ends of smooth rods under compression, provided a means of establishing that the commercial glass adopted by industry has a compressive strength of up to 200-250kgf/mm 2 with a small scatter of the data relating to properties. 3. Mechanical surface damage of commercial sive strength. LITERATURE I,
2. 3. 4. 5. 6. 7. 8.
1182
glass leads to a considerable reduction of its compres-
CITED
N. I. Kitaigorodskii (editor), Manual of Glass Production [in Russian], Vol. i~ GSI, Moscow (1965). O. A. Troitskii, "The strength of glass, " Priroda, No. 4 (1968). Catalogue of Engineering Microcrystalline Glasses [in Russian], Stroiizdat, Moscow (1969). G. N. Kuznetsov, Mechanical Properties of Rocks [in Russian], Ugletekhizdat, Moscow (1947). BL O. Gray and J. D. Stachiw, "Glass housing for hydrospace lights and instruments"~ Trans. ASME Series B, 92, No. 2 (1970). B. Kerley and R. Ravenhall, "interface stresses in ceramic metallic spherical shells", ALACk Paper No. 135 (1970). J. Herry and J. O. Autuoter, "Glass under the action of strong uniaxial compression, " in: The Strength of Glass [Russian translation], Mir, Moscow (1969). G. M. Okhrimenko and Yu. M. Rodichev, "Equipment for studying the strength and strain properties of brittle high-strength materials under biaxial compression conditions", Probl. Prochnosti, No. 1 (1973).