• No results found

a possible layer on the surface of the weld rod of either diffused additives or difficult to

using I. C.I Darvic clear 025 sheet, hereafter referred to as

range 5- 20mm?/min, see table 12 for details of the specimens used, A compliance calibration curve was only produced for a welded

2) a possible layer on the surface of the weld rod of either diffused additives or difficult to

remove dirt which acts as a physical barrier to fusion.

To make the weld strength less sensitiyh to the presence of notches in the weld zone, it is necessary to improve the toughs ness of the materials used to make the weld. It was considered impractical,in the short term,to change the composition of the sheet. However, changes to the weld rod formulation were

quite feasible and would allow the incorporation of impact modifiers into the weld rod, thereby improving the toughness of one of the

components cf. the weld.

The work carried out to explore both approaches to improving weld strengths was divided into two parts. In the first varied pre-weld treatments to the weld rod and sheet were used, whilst in the second, materials of different composition were obtained, converted into weld rod and used to produce hot gas welds, As such the second part was concerned not just with improving the toughness of the weld rod but also with trying to prevent the formation of any layer of additives on the surface of the weld rod.

All the welds prepared for this work were prepared as 300mm long symmetric double-vee butt welds with an included angle of

70°. Unless otherwise stated, each groove was filled with three passes of \?eld rod. Welding was carried out using

system B described in section 3.4.1. with, a gas jet temperature of approximately 280°C, measured 3mm from the end of the nozzle. Specimens were cut from the welded sheets and were prepared and tested as in.section 3.4,3. ^ 3,4,5,

3.6.1. Pre-Weld Treatment of Weld Rod

Hot gas \tfelds were prepared using clear weld rods which had been subjected to either an annealing treatment or a solvent . wash.

Six weld rods, 390mm long, were annealed for ten minutes at 150°C, on a bed of glass microspheres. After cooling in air, their lengths were found to be between 350-355mm, indicating that some relaxation of orientation had taken place. One weld

sheet was prepared with the annealed rod.

One of two solvents, either methanol or trichloroethylene, was used to wipe down both the surface of weld rods, and the

surfaces of the grooves and rods already welded in place, prior to each pass of weld rod. During welding, polyethylene gloves were worn to avoid re-contaminating the weld rod surface. Two welded sheets were prepared, one with rod washed with methanol and one with rod washed in trichloroethylene.

3.6.2. Changes to Weld Rod Composition

In conjunction with I.C.I. various changes to the standard

weld rod composition were decided upon, see table 13. Representatives of I.C.I. supplied the new compositions in the form of sheets

cut from the hide produced by a two-roll mixer. Approximately 300gms of each composition was supplied.

Weld rods of each composition were prepared by extruding the granulated sheets using a Davenport Rheometer' as a ram extruder. The sheets were first granulated such that the largest granules

were approximately 20mm . The rheometer was allowed to stabilise o

at 190- 5 C and approximately 70gm of granules were used to charge the barrel. The charge was added in lOgm lots which were tamped down after each, addition. Once the barrel was full, the ram was lowered until a slight pressure was read and the charge allowed to come to temperature oyer a 3-4 minute period. Ram extrusion was then carried out with the ram speed being adjusted to the fastest rate consistent with the production of smooth, bubble-free rod. Typical ram speeds were 10mmj?min, The

rod was allowed to extrude to a length of 350jnm before being

cut at the die. The extruded rods were then allowed to cool on a wooden bench top and were marked in the order they were extruded.

To extrude 3mm diameter rod it was necessary to manufacture new dies for the rheometer. The widest bore available in the

standard set of dies supplied by the manufacturer had an internal diameter of 2mm, During trials, using this die and granules of clear Darvic, the extruded rod was found to hqve a diameter of 2.5mm. It was calculated that a bore of 2,5mm internal

diameter was required, assuming constant die swell, to produce rod with a diameter of 3,1mm, Two stainless steel dies were produced, identical in size to the supplied dies except their bores had an internal diameter of 2.5mm. A H extrusion of weld rod of the new compositions was carried out using these dies,

It was normally possible to extrude 13^14 350mm lengths of rod from each 70gm charge. Although the barrel would take a larger charge, degradation of the polymer prevented any further useful rods being produced. The barrel and die were carefully cleaned after each run to prevent contamination of subsequent runs.

Only one welded sheet was prepared using rod of each

composition except in the case of formulation No.9, see table 13, for which a second welded sheet was prepared. This second sheet was welded using six passes of rod in each groove, which required two extra batches of rod to be extruded.

Selected samples of extruded rod were subjected to a heat treatment of 170°C for 10 minutes to determine the

level of frozen-in orientation caused by the extrusion process. The length of each rod, before and after the heat treatment, was measured. The rods were kept on a bed of glass microspheres during the heat treatment and the subsequent cooling.

Results

In the following chapter the results of the previously described experimental work are presented. The results are discussed in the context of the experiment from which they were obtained along with, their general significance to the overall programme of work. However, a more detailed

discussion relating specifically to the welding and fracture mechanics work is reserved for the next chapter.

Mechanical Properties

All the stresses reported here are engineering stresses-, and are calculated 1170] as the load, P, divided by' the specimen initial cross-sectional area, Aq,

a F/Ao -(12)

A true s t r e s s , , is defined as the load divided by the instantaneous cross-sectional area, A? and is related to the engineering stress, assuming a constant -volume deformation process, by

a - o x (

1/1

)

T ° -C13)

where 1 is the specimen gauge length under the applied load