SPECIFICATIONMethod of bondingThis invention relates to a method of bonding together two or more objects of cured epoxide resin and to articles made by this method.
Epoxide resins, by which is meant compounds containing more than one 1 2-epoxide group per average molecule, are well known materials, used for a wide variety of purposes, such as in castings and mouldings, as impregnants, and as adhesives. When two objects made of cured epoxide resin are to be bonded together, this is usually effected by applying an epoxide resin adhesive formulation to the surfaces to be joined and maintaining the objects in the requisite position whilst the adhesive cures. By using an epoxide resin adhesive the known advantages of an all epoxide resin system, such as good electrical insulation and chemical resistance, are retained and the bond has a high strength.The curing process for the adhesive is, however, often inconveniently slow, and whilst it may be accelerated by heating, this is not always convenient, due to the size or location of the objects being bonded.
There is therefore a need for a rapid method of bonding together objects made from cured epoxide resins without introducing other, epoxide-free materials. We have now found that this can be achieved by means of friction bonding under certain conditions, whereby frictional heat caused by moving one object relative to the other whilst they are held in close contact causes the epoxide resin composition in the area of contact to attain a soft, rubber-like state, this movement is then stopped, and the contacting surfaces, on cooling, form a strong bond between themselves. Full bond strength is reached in a matter of seconds, rather than of hours as is usual with a conventional epoxide resin adhesive.
This bonding technique bears a superficial similarity to friction welding, which is well known for bonding metal or thermoplastics articles. However, such materials have definite melting points and fusion occurs between the contacting surfaces. Thermoset resins, such as cured epoxide resins, on the other hand, do not melt but only soften when heated and, if heated further, they finally decompose without melting. It was therefore not to be expected that bonding of objects of cured epoxide resins could take place through the generation of frictional heat.We have further found that, provided the cured epoxide resins have a glass transition temperature within the range 400 to about 2000 C, successful friction bonding can take place; cured resins having a glass transition temperature below this range can be friction-bonded, but the resultant joints are usually too weak for practical use, while resins having a glass transition temperature above this range tend to undergo, at the requisite bonding temperature, decomposition at the interface to an extent such that the joint strength is below acceptable levels.
Accordingly, this invention provides a method of joining two objects made of cured epoxide resin, the cured epoxide resin of each object having a glass transition temperature within the range 400 to 2000C, and preferably for at least one such object within the range 800 to 1 C, comprising::a. placing the two objects in contact with each other at substantially planar mating surfaces,b. applying a static force to the said objects to urge them into intimate contact with one another at said surfaces,c. subjecting said surfaces to motion, relative to one another, for a sufficient period for the frictional energy between the said surfaces to heat the said surfaces to above the glass transition temperature or temperatures of the objects, andd. stopping, or allowing to stop, the said relative motion whilst maintaining the said static force until the mating surfaces have cooled to below the glass transition temperature or temperatures of the objects.
This invention further provides articles made by this method.
By the term 'glass transition temperature, as used in this specification and claims, is meant the approximate midpoint of the temperature range over which the cured epoxide resin changes reversibly from a hard and brittle condition to a rubbery condition, as measured by differential scanning calorimetry, torsion pendulum, thermomechanical analysis, or other acceptable means.
The relative motion between the objects may be reciprocative, with one or more objects being vibrated back and forth with respect to another in the plane of the mating surfaces, or, preferably, the motion is rotational, with one or more objects being rotated against another.
When a rotational method is employed with objects held in contact end to end, the relative speeds of the surfaces in contact, and hence the frictional heating effect, reduce to zero on the axis. We have found that better results are achieved if a small area around the axis is removed from one or both mating surfaces. The area of surface removed usually represents 25% or less of the total mating surface.
The objects to be bonded will usually be rods or tubes.
The bonding of two coaxial rods end-to-end, the relative movement between them being rotational, is illustrated in Figure 1 of the accompanying drawings.
Overlap joints may be produced by rotation of a closely-fitting sleeve or ring of cured epoxide resin around cylindrical objects, also of cured epoxide resin, both of which may be stationary or one of which may be rotating relative to the other. The two rods may abut one another, or they may almost meet, and advantageously the rods are tapered towards the contiguous or adjacent faces while the sleeve or ring tapers outwards. This arrangement is illustrated in Figure 2 of the accompanying drawings.
Alternatively more than two surfaces can be bonded simultaneously by arranging co-axially three objects of cured epoxide resin, each face of the central object contacting the planar mating surface on a face of the adjacent object, and causing the central object to rotate, such as by means of a rotating belt, or vibrate, relative to the other two objects while applying a static force to the three objects to urge them into intimate contact with one another at the said surfaces. Such an arrangement is illustrated inFigure 3 of the accompanying drawings.
During the bonding process the cured epoxide resin tends slightly to bulge outwards in the immediate area of the bond, due to the heated resin yielding to the force applied to hold the mating surfaces together. Examination under polarised light of the loints so produced shows that the bulging "flash" of the epoxide resin is more highly stressed than the remainder of the joint, and so it is sometimes found advantageous to machine away this flash, particularly if the joint is liable to be heated in service to a temperature exceeding the glass transition temperature of either object.
Any conventional epoxide resin may be employed to make the objects which are bonded by the method of this invention. Preferably they contain at least one group of formula
directly attached to an atom of oxygen, nitrogen, or sulphur, where either R and R2 each represent a hydrogen atom, in which case R' denotes a hydrogen atom or a methyl group, or R and R2 together represent -CH2CH2-, in which case R1 denotes a hydrogen atom.
Particularly preferred is an epoxide resin which is a polyglycidyl ether of 2,2-bis(4hydroxyphenyl)propane, of bis(4-hydroxyphenyl)methane, or of a novolak formed from formaldehyde and either phenol or phenol substituted in the ring by one chlorine atom or one alkyl hydrocarbon group containing one to nine carbon atoms, and which has a 1 2-epoxide group content of at least 0.5 equivalent per kilogram.
The epoxide resins employed to form the objects may be cured by any conventional curing agent for epoxide resins, such as aliphatic, cycloaliphatic, aromatic, or heterocyclic polyamines, polyaminoamides, adducts of polyamines with stoichiometric deficits of polyepoxides, isocyanates, isothiocyanates, polyhydric phenols, phosphoric acid, polythiols, and polycarboxylic acids and their anhydrides. There may also be used catalytic curing agents such as tertiary amines, alkali metal alkoxides, stannous salts of alkanoic acids, dicyandiamide, Friedel-Crafts catalysts, and chelates formed by reaction of boron trifluoride with, e.g., 1 ,3-diketones.
An object made of an epoxide resin composition may be bonded to one made of a different epoxide resin composition, i.e., one containing a different epoxide resin, or a different curing agent, or a significant amount of a flexibiliser, for example. However, if the best results are to be achieved, the glass transition temperatures of the two objects to be bonded together should not differ by more than about 50 C.
The compositions forming the objects may contain inert fillers, reinforcing materials, and colouring matter, if desired. It is, however, preferred that the total volume of such additives constitute at most 35 parts by volume per 100 parts by volume of the cured epoxide resin, although larger proportions may be used. A total of at most 20 parts by volume of such additives per 100 parts by volume of the cured resin is further preferred. Bonding of objecis in which at least one of them is substantially free from such additives is particularly preferred.
The invention will now be illustrated by the following Examples, in which all parts are by weight and temperatures are in degrees Celsius. The resins and hardeners used in the Examples were as follows:EPOXIDE RESIN Ithis term denotes a diglycidyl ether of 2,2-bis(4-hydroxyphenyl)propane, having an epoxide content of 5.0-5.2 equivalents/kg.
EPOXIDE RESIN IIthis denotes a solid resin prepared from 2,2-bis(4-hydroxyphenyl)propane and epichlorohydrin and having an epoxide content of 2.2-2.5 equivalents/kg.
HARDENER Ithis denotes an adduct of an equimolar amount of triethylenetetramine and propylene oxide.
HARDENER II this denotes triethylenetetramine.
HARDENER Illthis denotes 4,4'-diaminodiphenylmethane.
HARDENER IVthis denotes phthalic anhydride.
EXAMPLE 1Rods of circular cross section 1 5.7 mm in diameter were cast from Epoxide resin I and various hardeners. The nature of the hardeners, the amounts added, the curing conditions, and the glass transition temperatures (Tg) of the cured samples (measured by differential scanning calorimetry), are given in Table 1.
TABLE 1
Hardener Parts per 100 parts Sample Designation of Epoxide resin I Curing conditions Tg A I 25 24 hours at 20" 93" B 11 12 24 hours at 20" and 3 hours at 70" 95" C 11 12 24 hours at 20" and 3 hours at 95" 116" D i i 12 3 hours at 60" and 119" 3 hours at 95" E Ill 27 16 hours at 60" and 3 hours at 95" 137" F Ill 27 16 hours at 60" and 3 3 hours at 95" and 3 hours at 120" 142" G lil 27 16 hours at 60" and 3 hours at 95" and 3 hours at 120" and 3 hours at 150" 168" H Ill 27 16 hours at 60 and 3 hours at 95" and 3 hours at 120" and 3 hours at 150" and 3 hours at 180" 171" The rod specimens were prepared by first lightly machining a flat face on one end of each rod, then counterboring these surfaces to a depth of 2 mm and a diameter of 7.5 mm. One rod was then mounted in the three-jaw self-centering chuck of a lathe and rotated at 1280 rpm. A second, similar rod was held in a chuck mounted in the tailstock of the lathe so that it could not turn but could be moved axially. This second rod was then forced against the rotating rod, pressure being applied by means of the tailstock handwheel via a calibrated compression spring device, so arranged that the axial thrust force between the two resin rods was known and reproducible.After a short period of rubbing contact between the rods as recorded in Table 2, the lathe drive was disengaged and a brake applied to bring the chuck to rest as quickly as possible, axial pressure being maintained for a further half-minute whilst the specimen cooled. The bonded specimen was then removed in one piece.
The shear strength of the bond was found by twisting one end of the article and measuring the torque on the other end at break. The results are given in Table 2, each result being the average of three replicates.
TABLE 2
Bonding conditions Time of rubbing Shear strength Sample Force (N) contact (secs) (MN/m2) A 300 7 48.1 B 300 8 49.9 C 300 9 50.9 D 330 9 46.4 D 360 9 46.5 E 330 9 40.9 E 360 9 54.6 F 360 9 54.3 F 360 10 48.1 G 360 10 62.5 G 360 11 55.8 H 360 11 61.3 H 360 12 62.3 EXAMPLE 2Example 1 was repeated, the rods being made of 100 parts of Epoxide resin I and 27 parts ofHardener Ill, curing at 1000 for 3 days. The rods were then bonded together as described in Example 1, using an axial force of 360 N for 9 seconds, the 'flash' around the bond area was removed by machining the article to an external diameter of 12 mm. The article was post-cured at 1800 for 20 minutes. The shear strength at room temperature, determined as in Example 1, was found to average (8 replicates) 56.8 MN/m2.
EXAMPLE 3Example 1 was repeated, rods being made either from 1 00 parts of Epoxide resin I and 27 parts ofHardener Ill, curing at 650 for 16 hours plus 2 hours at 1800, or from 100 parts of Epoxide resin II and 30 parts of Hardener IV, curing for 16 hours at 1 350 The bonds were formed as described in Example 1, using an axial force of 360 N for 9 seconds. The tensile strengths were measured by pulling the bonded rods in an Instron testing machine and recording the force required to separate the joint. The average tensile strengths of the bonds (4 replicates) were 8.8 MN/m2 for the rods made from Epoxide resin I and 7.5 MN/m2 for those made from Epoxide resin II.
EXAMPLE 4Example 1 was repeated, bonding rods made from 100 parts of Epoxide resin I and 27 parts ofHardener Ill and cured at 1400 for 1 hour to rods made from 100 parts of Epoxide resin it, 30 parts ofHardener IV and 200 parts of silica four, cured at 1350 for 16 hours.
Bonding was effected as described in Example 1, using an axial force of 360 N for 7 seconds.The shear strength of the bond, measured as described in Example 1. was 13.7 MN/m2.