It should be emphasised that in this specification, including the claims, the meaning of the term "blowing agent" encompasses both "chemical blowing agents" and "physical blowing agents" A chemical blowing agent comprises at least one compound which breaks down in molten plastics material to mix gas into the material.

A physical blowing agent is a gas or a liquid - at room temperature and pressure - or mixed with the molten material. Chemical blowing agents are added to the plastics material in the hopper of matel ial fed to an injection moulding machine's plasticising screw. Physical blowing agents can be mixed with the plastics material as described in my UK patent applications Nos. 97029771.1 and 9706682.3, respectively dated 13th February and 2nd April 1997. These are still unpublished. However, the mechanism for producing a mixture of blowing agent and plastics material forms no part of this invention which is concerned with production of articles using such mixtures.

Generally, the blowing agent as such will acld negligibly to the volume of the material when being processed within an injection moulding machine due to the elevated pressure within the machine. However, when the pressure is released to substantially lower pressure, such as room pressure, and the plastics material is still molten, the blowing agent acts to expand the material by the blowing voids of varying sizes in the material.

A recognised problem. in section moulding fiom plastics materials of articles having thin walls in particular, is that at marked differences in cross-section, differential shrinkage causes blemishes in the finished surface of the article.

Despite this problem, many new designs of injection moulded articles could be feasible if marked changes in cross-section were possible, without shrinkage blemishes.

' Cllcmic.ll blowing agellls are llsllall! botiiid ijito a c;irrier oSotiler plastics material xvllich has its own bulk.

The following applications describe my basic invention having the object of providing an improved method based on injection moulding and facilitating the production of substantial changes in cross-section: British Patent Application No. 95 14674.2 US Patent Application No. 60/017386 International Patent Application No. PCT/GB96/01706, now published under No WO 97/03800.

In these applications 1 describe and/or claim (as proposed to be amended): A method of forming an article via injection of plastics material into a mould, the finish formed article having thin wall portion(s) and thick wall portion(s), the thick wall portion(s) being at least partially foamed, the method consisting in the steps of: providing a mould tool defining in its closed state, between its cavity part and its core part, narrow gap portion(s) whose mould part gap is to be substantially reproduced in the thin wall portion(s) of the article and wide gap portion(s) whose mould part gap is less than the thickness of the thick wall portion(s) of the finish formed article, closing the mould tool to deLhle the narrow and wide gap portions; injecting a plastics nlate ial mixture comprising a basic polymer and a foam producing additive into the mould tool; allowing the plastics material mixture to at least substantially solidify in the narrow gap portions of the mould tool to produce the thin wall portions of the finish formed article; withdrawing at least a portion of one part of the mould tool from the other part before the plastics material mixture has at least substantially solidified in the wide gap portion(s) of the mould tool to allow the mixture to expand by foaming and form at least some of the thick wall portion(s) of the finish formed article; and ejecting the article from the mould tool.

This method is hereinafter referred to as "My Original Method"

To help understanding of My Original Method its first embodiment will now be described with reference to the following of the accompanying drawings, in which: Figure 1 is a cross-sectional side view of an injection moulded cup able to be moulded with a mould using a technique, which appears to be known; Figure 2 is a similar view of a cup formed in accordance with My Original Method; Figure 3 is a similar view of a mould tool for preliminary moulding of the cup; Figure 4 is a side similar view of the mould tool open for foaming of thick wall sections of the cup on the core ofthe mould tool.

Referring first to Figure 1 5 the cup has a base i, a lower side wall 2 and an upper side wall 3, these being thin wall portions. At the corner 4 between the base and the lower side wall, at a band 5 between the upper and lower side walls and at the rim 6 there a thick wall portions. Typically the thin wall portions are 0.7mm thick and the thick wall portions are 1.2mum thick.

This cup could be moulded with conventional plastics materials, but due to the different wall sections, shrinkage marks could be expected to appear in the thick wall portions. In other Words, conveiitional moulding techniques result in uneven wall thickness in the thick wall poltio ls I discovered that I can mould the cup with even wall thickness in the thick wall portions, by including a small amount of chemical blowing agent in the plastics material. Despite having originally believed this to be a new technique, it appears that this may be known.

In this basic technique, I used plastics material comprising free flowing polypropylene with a small addition of blowing agent, typically less than 5% and in the region of 1%, in accordance with the directions of the suppliers of the SAFOAM agent, Reedy International Corporation of Keyport, NJ, USA. In the thick wall portions, the agent causes foaming, whilst in the thin wall portions no foaming occurs.

The degree of foaming can be controlled by adjustment of injection parameters such as pressure, time, temperature, quantity of plastics material and percentage of blowing

agent in the material, such adjustments being routinely made in the set up of an injection moulding machine.

I believe that a combination of the higher pressure required to force the material into the thin wall portions and the increased cooling rate in the thin wall portions inhibit the formation of foaming in the thin wall portions whereas the lower pressure present in the thick wall portions and the greater bulk of plastics material in the thick wall portions requiring longer to cool allow foaming in these portions.

Originally I believed that the foaming to fill the thick wall portions of the mould tool cavity needed to occur before opening of the mould. However I was surprised to discover that additional foaming can occur due to opening of the mould before cooling ofthe thick wall portions to solidification.

My Original Method is based on such use of plastics material including blowing agent to allow foaming expansion to continue after at least partial opening of the mould in which the cup or other article is moulded.

Figure 2 shows a cup formed in accordance with My Original Method. It has thin wall portions, namely a base 101, a lower side wall 102 and an upper side wall 103, in which no foaming occurs. These portions have their wall thickness determined by the mould part gap. The cup also has thick wail portions, namely the corner 104, the band 105 and the riln 106, in which foaming occurs after mould opening so that the wall thickness is increased beyond that provided by the mould. Compared with the 1.2mm wall thickness in the band 5, using the same mould, a maximum band wall thickness of 3.2mm is achievable. It will be noted the outer contour of the band is curved, due to restraint of its upper and lower margins 106,107 where the wall thickness alters to being thin. On the outer surface 108 between these margins, the band bows out. The outer surface of the rim 106 also bows out. In both these instances, and indeed at the corner 104, the outer surface is substantially solidified on opening of the mould. At the time of the above referenced patent applications, I believed that the outer surface was able to stretch as foaming occurs in the still molten plastics material at the centre of the thick wall portions, to give the shapes shown in Figure 2. In the cup of Figure 1, such foaming as occurs, does so against the

constraint of the still closed mould. In the cup of Figure 2, the constraint on the foaming is atmospheric pressure and the skin tension of the outer surface. The shape of the rim outer surface 109 is of particular note, in that the shape in which it originally solidifies is concave. However on foaming, the concave surface has little pressure constraint on it and is blown out over-centre to the convex shape 109. This results in a maximum wall thickness of 2.7mm despite the vertical extent of the rim being less than that of the band.

It should be noted that the 3:1 thick:thin wall thickness ratio, that is the ratio of thin wall thickness to thick wall thickness prior to foaming after mould opening, is exemplary only and the possible limits on the ratio have not yet been researched.

Turning now to Figure 3, the mould tool there shown has a cavity 11 and a core 12, the two being separable at ajoint line 13. The cavity has an injection point 14 and a spring closed air injection port 15 in the form of a poppet valve. The core has an air injection port 16 opposite the injection port 15, with a free floating valve member closed by pressure in the moulding void.

On injection of the plastics material mixture, the port 16 closes and the moulding void fills. The injection parameters are adjusted such that the material reaches the cup rim 6, without completely filling the void. The blowing agent causes foaming in the thick wall portions. However this does not occur in the thin wall portions, where the pressure required to displace the material is higher and the cooling is quicker. When time has been allowed for the plastics material to at least substantially solidify in tile thin wall portions ofthe mould 1',2',3', corresponding to the base 1, lower wall 2 and upper wall 3 of the cup and for some foaming in the thick wall portions 4',5',6' of the mould, corresponding to the corner 4, band 5 and rim 6, and before the material has solidified in the thick wall portions, the mould is opened and air pressure applied to the port 15. This separates the moulding from the cavity, and together with shrinkage onto the core, allows withdrawal of the core with the moulding.

The outer surface skins 108, 109, 10 of the material (see Figure 2) at the thick wall portions 104, 105, 106 have at least substantially solidified, but does not constrain the blowing agent from generating sufficient pressure to create the shapes described above with reference to Figure 2 and as also shown in Figure 4. After a further delay to allow the moulding to cool, air is introduced via the port 16. This expands the moulding which is released from and then drops off the core.

Figure 4 is taken from my International Application No. PCT/GB96/01706. At the time ofthe application, I had not realised the significance of the present invention.

For this reason, Figure 4 shows the rim 106 abutting the joint face surface of the core part of the tool. When sufficient blowing agent is added to the material of the cup and the cooling time is sufficiently short, for the rim to form to a circular cross-sectional shape, the height of the cup fore-shor-tens due to the rim skin blowing out and down as its length remains constant. Thus Figure 4 does not teach the present invention.

I believe that the blowing agent forces the moulding into good thermal contact with the mould at the thick wall portions before opening of the mould. This enhances cooling of the plastics material to form the skins of the thick wall portions. Further use of a carbon dioxide blowing agent which absorbs appreciable energy in foaming, that it cools the material on foaming, is advantageolls in quickening cooling. However, both chemical and physical blowing agents using > other gases and base polymers other than polypropylene are possible to use in My Original Method.

In British Patent Application No. 9621626.2, from which this application claims priority, I described an improvement of My Original Method.

In the improvement, a substrate is placed into the mould prior to its closure in a position such that tlie plastics material mixture is injected onto a surface of the substrate, whereby after the withdrawal step the mixture foams on the substrate.

This improvement is hereinafter referred to as "My Substrate improvement".

I have now realised that the foaming expansion mechanism described in respect of My Original Method is not exactly as I had originally believed, particularly as regards the extent to which the skin in the foamed thick wall portions stretches. The degree of stretch now appears to be very small and limited by solidification of the skin very shortly after the foaming pressure is applied to it. This effect is apparent with My Substrate Improvement where the lack of stretch, but change of cross-sectional shape pulls the substrate into a curve.

In application No. 9621626.2, 1 anticipated the applicability of My Substrate Improvement to articles having no thin wall portions.

To help understanding of My Substrate Improvement, its preferred embodiment as described in the application No. 9621626.2 will now be described with reference to Figures 5, 6 Kr 7 ofthe accompanying drawings, in which: Figure 5 is a cross-sectional side view ofa mould tool having a substrate placed in it prior to injection of plastics material, Figure 6 is a similar cross-sectional view of a formed article after foaming of the plastics material on a surface of the substrate and Figure 7 is a plan view of the formed article.

The mould tool has a cavity part 211 and a core part 212, defining a cuboid cavity 213 having a thin wall section 214, which extends over the major part of the square plan of the cavity, and a thick wall section 215 which extends along one edge of the square. This tool will be understood to be intended for manufacturing samples.

The square has 80mm sides. The thin wall sections are 1.2mm thick and thick wall section is 4.2mm thick, that is the core part ofthe mould has groove 216 3mm deep and 10mm wide along one side of the square.

In formation ofa sample 221, an 80mm square ot0.2mm thick, polyethylene faced card 222 was inserted into the mould, with its face away from the cavity part.

Foamable plastics mixture was injected into the cavity via a port 217 in the core part.

The card was forced against the core part 211, so that the plastics material formed a

lmm thick skin 223 across the back of card except at the one edge 224 where the plastics material had an initial thickness of 4mm.

On opening of the mould, the sample was ejected and the plastics material in the edge allowed to foam to form a ridge 225, whose thickness increased to 9mm. In so doing it caused the card to adopt a local curvature 226. The plastics thickness across the rest of the card remained l.Omm, since the plastics material here had solidified prior to mould opening.

Although not described in application No 9621626.2, initially I had formed the thick wall section groove 216 with a half round cross-sectional shape. This did not foam to such a marked extent as I expected. It was when the shape was changed to be rectangular that the degree of foaming described was achieved. It should be particularly noted that the degree of foaming shown in Figure 7 is such that the cross- sectional shape of the ridge 225 is nearly circular and the card has taken up the local curvature 226 tending towards the circular cross-section.

I now believe this to result from the circumferential length of the skin around the thick wall section restraining the degree of foaming expansion. The circumferential length around a rectangle is of cour-se greater than that of an inscribed circle.

The limitation on foaniing expansion to the circumferential extent in transverse cross-section of the region of the mould fi-om which the foamed portion of the article is formed renders the process surprisingly predictable.

Therefore, when employing My Original Method, I now prefer to foam the thick wall portions, or the entire object where appropriate, to a substantially circular cross-section, at least where a high stitfness is required (and other constraints do not call for alternative shapes). However, I should emphasise that the expansion to substantially circular cross-section is applicable beyond the differential wall thickness articles of My Original Method.

Thus in accordance with the present invention, I provide a method of forming an article via injection of plastics material into a mould, the finish formed article having at least one portion expanded, by blowing agent, to a shape which in cross-section is at least substantially circular over at least a substantial part of its circumference, the circumference having a predetermined extent, the method consisting in the steps of: providing a mould tool defining in its closed state, between opposite parts, at least one region having a circumferential extent around the region in transverse cross- section substantially equal to the predetermined circumference; closing the mould and injecting a plastics material mixture comprising a basic polymer and a blowing agent into the mould tool; allowing the plastics material to skin adjacent thereof with an extent substantially equal to the predetermined circumference; withdrawing at least a portion of one part of the mould tool from the other part before the plastics material mixture has at least substantially solidified at the said region(s) of the mould tool to allow the mixture injected therein to expand by action of the blowing agent to the said at least partially, at least substantially circular shape and form the said portion(s) of the finish formed article; and ejecting the article from the mould tool.

I refer to the article as skinned, but before expansion on mould withdrawal as a "pre-form" Thus my present invention includes My Substrate Improvement when expansion occurs to the said circular shape.

Whilst 1 can conceive that the expansion may occur after ejection of the article from the mould tool, normally the expansion will occur before this.

Preferably, the foaming expansion is to an extent whereby the substantially even radius of curvature of the skin is present over a major arc if not the entire circumference of the said portion(s).

Preferably the or each said region is wider than it is thick. In one alternative, the withdrawal step is delayed from a point in the cycle time when the expansion would be to an extent whereby a substantially even radius of curvature of the skin is present over substantially the entire circumference of the article in the said portion(s), whereby the skin stiffens to such extent that edges of the expanded portion substantially reproduce the corresponding edges of the mould region(s), whilst the intervening sides are bowed to at least partial, at least substantially circular shape, the said portion(s) tending to elliptical shape.

In another alternative the expansion occurs against a mould tool surface, whereby the said portion of the finish formed article includes a face from which the transverse section of the portion extends with partially circular curvature.

It should be noted that the said portion(s) may comprise the entire article.

Preferably the mould part gap at the said region(s) will be at least partially elongate, having a constant cross-section.

The said portion(s) may have a thin wall portion at one end. Additionally or alternatively, the poltion(s) may have a thin wall portion at one side.

The said region(s) of the mould may have a cross-sectional shape having one or more marked discontinuities of contour, such as a relatively tight radius of curvature, whereby the discontinuities provide a skin circLln7fel-entiaI length which can be extended with a substaiitially even and greater radius of curvature on foaming expansion. Examples of such shapes are the rectangular shape of the groove of Figure 5 above and tlie tear-drop shape described below. It should be noted that shapes having an appreciable difference in transverse dimension taken normally to each other can be expected to fore-shor-ten in their longer dimension on foaming expansion.

Alternatively, or additionally, the said region(s) may have a cross-section shape having at least one marked concavity, whereby the corresponding skin concavity can be inverted to provide at least one characteristic dimension of the cross-section of the

portion which substantially increases on foaming expansion. Examples of such shapes are alphabetic shapes having marked concavities, such as C, H, I, M/W, UIV, X, Y.

Where a shape having a sharp corner is used, such as a rectangle or an X, the sharp corners are preferably slightly relieved to avoid vestigial traces in the final skin shape.

The or each discontinuity is a tight radius of curvature which is tight compared with that of the transverse cross-section of the corresponding portion of the finish formed article.

Further, the discontinuity can be at a feature of the region(s) causing locally accentuated cooling whereby the feature is locally frozen into the skin at the at least partially, at least substantially circular shape of the said portion of the finish formed article.

Shapes having more thaii one concavity and a central part of the portion which can act as a reservoir of liquid loaming material provide the greatest scope for increase in overall cross-sectioiial dimension.

In some embodiments, the or each said region is curved - with a local radius of curvature which is large in comparison with the dimensions of its cross-section - and defines a flange extending radially and at least one laterally facing concavity, whereby the concavity inverts substantially laterally on expansion. Preferably, the or each region defines a web from which the flange extends, the web remaining in cooling contact with one of the mould parts during the expansion. The or each region can include a radially extending rib defining a groove in the or each web on expansion, the rib remaining in cooling contact with its groove.

Where the substantially circular cross-section foamed portion adjoins a thin wall portion, transitional curvature will result. If the foamed portion is biased to be to one side of the thin section, by for instance allowing the foaming expansion to occur

with the thin section still in contact with a mould part, the foamed portion will have asymmetric transitional curvature joining opposite sides of it to the thin wall portion.

The method of my present invention is particularly applicable to the production of cutlery. In the preferred embodiments described below, thick wall/foamed portions extend along edges of the handles of the cutlery. Preferably, the thick edge portions have thin portions between them, at least in the handles. The thick portions may extend around the distal end of the handle. Alternatively, the thin, central portions may extend to the distal end of the handles. One or more further thick portions may extend longitudinally of the handle between the edge thick portions.

The thick portions. particularly the edge thick portions, may extend into the "cutlery" part of the cutlery. By "cutlery" part is meant the blade of a knife or prongs of a fork or bowl of a spoon. Further reinforcing thick portions may be provided at the "swan's neck" of a fork or spoon. By "swan's neck" is meant the transition between the cutlery part and the handle. Further, the prongs of a fork may be formed as thick/foamed portions. However, it is likely that other portions of the fork will be thicker/foamed to a greater extent.

In a recent developinent of the method of my present invention, I have now produced a series of articles, wherein the said portion(s) are foamed to such extent that a void forms at the centre of the portion. The article is in effect hollow in the portion(s).

Where the portion - or one of them in the article - is to be handle, this is a particularly advantageous feature, in that it creates a handle having a body which can be readily gripped. Further since the handle has a considerable moment of inertia, it is stiff and strong.

I have also produced annular hollow portions in the same manner. These also can have considerable stitftess and strength.

If I use a soft plastics material, the hollow portion becomes readily compressible yet resilient due not only to the resilience of the material but also to the gas enclosed within the hollow. I envisage the portion being used as a seal. For instance, a small container with such a hollow portion around its rim could be an eye bath.

Further, I have produced a series of articles in which three phases are present, namely thick and thin portions in accordance with my Original Method, in which the thin portions are unexpanded and the thick portions are expanded but not hollow, and thicker portions still where the opposite skins have separated to leave a hollow, central void. I refer to these phases as: i. Solid, ii. Cellular, iii. Hollow.

There are a number of variables which determine whether a particular portion will become solid, cellular or hollow. For a given material, the determinant of solidity is whether the material has been cooled sufficiently to be solid throughout when the mould is open. No invariable rules can be given for this, nor indeed is it easy to make measurements of solidity on mould opening. It can be deduced by the finished article's wall thickness substantially reproducing the mould part gap. Cooling is enhanced by running the mould as cool as possible and delaying mould opening after injection for sufficient time for solidification. Excess cooling and/or delay in mould opening should be avoided to ensure that foaming expansion on mould opening can occur in the non- solid portion(s).

The chief determinant of a cellular phase is that the cross-sectional shape of the region should be such as to allow the skin to change shape on mould opening, but to a limited extent only. For instance, if the initial skin shape - i.e. the shape of the skin as moulded before foaming expansion - is convex, there is less scope for its inflation to 2 l use the terill "cclltitar" iii colllrast to..fo;llllcd". because botll the "cellul;tr'' and "hoilow" phases result from a process of foaiiiug.

permit expansion of the cross-section. An extreme example is of a circular cross- section initial skin shape. This can inflate only by stretching of the skin, whereby even if the central portion of the plastics material is still molten on mould opening, a small amount only of foaming expansion can occur. Another example is of an equilateral cross-sectional shape. This can inflate by bowing of the sides and moving inwards of the corners, whereby foaming expansion can occur and hence a cellular structure can result.

The chief determinant of a hollow phase is again cross-sectional shape of the region. Creation of a hollow phase requires that the skin should be able to change shape to a much greater extent, such that the cellular structure expands so much that the individual cells burst. For instance, a rectangular shape is likely to expand on foaming expansion to the extent of becoming hollow. The likelihood of this increases with the aspect ratio - i.e. the ratio of the longer sides to the shorter sides - of the rectangle. Again, concavity in the initial skin shape is likely to produce an inflation sufficient to allow foaming expansion to the hollow phase to occur.

A factor influencing the degree to which the skin will inflate is the composition of the plastics material. Materials which will stretch to a greater extent before being inhibited from filrther stretching by ciystallinity and/or biaxial orientation will allow the hollow phase to develop to a greater extent. However, my experience to date in working primarily with polypropylene materials is that cross-sectional shape is a more important design criterion than material nature as regards the creation of hollow or cellular phases.

Whilst the basis of my development has been to produce articles or parts of articles having substantially circular cross-section, I can envisage uses where a flattened circular cross-section would be useful. Particularly, where a greater moment of inertia in one plane than an orthogonal one is required, as for instance in knife handle. To produce such a shape, 1 envisage partially opening the mould and allowing the expansion to occur against the mould surfaces. Alternatively, it is possible to produce cross-sectional shapes, which whilst tending towards circular are intermediate between the initial skin shape and a circular cross-section. For instance if 5% blowing

agent and l/2 second cooling time is required to expand a rectangular initial shape to circular, 3% blowing agent and 3/4 second cooling time is likely to result in insufficient fluidity in the centre of the moulding and flexibility in the walls for the cross-section to expand to being fully circular. The long sides of the rectangle may bow out to have a substantially even radius of curvature. However the constraints from the short sides and the corners between the sides may be such that the cross-sectional shape is tending to elliptical, or indeed some more complex mathematical shape. At this stage, I have not had the opportunity to analyse the exact curvature.

However, l believe the shape to be tending to be substantially circular or at least substantially elliptical, because if sufficient blowing agent - or other foam producing additive - is included in the plastics material and the mould is opened as soon as the skin has formed then a substantially circular cross-section will result if the initial shape is sudi as to allow this.

I therefore think in terms of this part of the development having been taken as far as the foaming producing an expanded cross-section having convex curvature tending towards circular, which is absent in the initial skin shape, and a hollow phase within the skin.

I would eiphasise tliai the skin surrounding a hollow phase is likely to be at least partially cellular, with the cell size increasing towards the hollow phase.

There are two factors which appear to have an effect on the extent to which a hollow phase forms. Firstly, I believe that in certain circumstances heat soak has an influence. By heat soak, l mean that the moulding is sufficiently rigid - due to cooling of the skin - on mould opening that expansion is not immediate. Then after a perceptible delay, the moulding grows in size. This appears to be caused by some of the heat in the interior of the moulding, where the material is still molten or at least readily extensible, being transferred out to and warming the skin, softening it and/or the melt layer thickening. The softening/thickening reaches a point where the skin can no longer resist the expansive. internal pressure and the hollow or cellular phase(s) form. This softening is ofa diffi se nature. It does not necessarily result in certain

moulded features disappearing fi-om the tinal article. Heat soak can be utilised to encourage expansion to initiate at a specific points, by increasing the cross-section, typically centrally. Once expansion has been initiated it will tend to spread across the moulding.

The second factor is the converse of heat soak, namely the extent to which the moulding is cooled. More cooling will tend to occur at the edges of a moulding having an appreciable aspect ratio. Accentuating cooling can cause features to be "frozen into" the final article. Certain features will tend to be cooled to a greater extent due to tapering to an edge. Such features are more likely to be frozen in.

Within the scope of mv development is that I can produce articles having no solid nor cellular phases except in the skin. I can also produce articles which have solid and hollow phases, but no separate skinned cellular phases without hollow interioi-s. My nile ofthumb is that if the degree of expansion - i.e. the increase in transverse dimension - exceeds 5mix, a void will form.

My present invention can be used for forming articles which do not have respective concave and convex sides, but rather are symmetrical about a central plane, for instance a knife. It niay be thought that the mould tool for such articles is not properly to be regarded as having a core part and a cavity part, but rather has two opposite complementar-y mould parts. I believe that the man skilled in the art will refer to such a mould as having core and cavity parts by analogy with conventional usage.

Nevertheless, for the avoidance of doubt as to the scope of the present invention and My Original Method, both cover the use of symmetrical moulds, having mirror image mould cavity defining parts.

Use of such a mould is not described in my International Application No.

PCT/GB96/0 1706, which was published on 6'1' February 1997. Such use was described in Applications Nos. 9624162.5 and 9700138.2, dated 20"'November 1996 and 6th January 1997 respectively, from which this application claims priority. Thus a claim to such use is valid in this application and is included in the claims hereof

Thus according to this aspect of the invention there is provided: A method offorming an article via injection of plastics material into a mould, the finish formed article having thin wall portion(s) and thick wall portion(s), the thick wall portion(s) being at least partially foamed, the method consisting in the steps of: providing a mould tool defining in its closed state, between its two opposite parts, narrow gap portion(s) whose mould part gap is to be substantially reproduced in the thin wall portion(s) of the article and wide gap portion(s) whose mould part gap is less than the thickness of the thick wall poltion(s) of the finish formed article, the two opposite mould parts being symmetrical; closing the mould tool to define the narrow and wide gap portions; injecting a plastics material mixture comprising a basic polymer and a foam producing additive into the mould tool; allowing the plastics material mixture to at least substantially solidify in the narrow gap portions of the mould tool to reproduce the thin wall portions of the finish formed article; withdrawing at least a portion of one part of the mould tool from the other part before the plastics material mixture has at least substantially solidified in the wide gap portion(s) of the mould tool to allow the mixture to expand by foaming and form at least some of the thick wall portions of the finish formed article; and ejecting the article fi-oni tlie mould tool.

Further, it may be thought that an article such as a piece of cutlery as opposed to a container is not properly to be regarded as having a thick "wall" portion and a thin "wall" portion. I believe that the man skilled in the art will realise that the term "wall thickness" means the "gauge thickness" (a common term for the thickness of material) or indeed merely the "thickness" of the material. both as defined by the mould part gap and measured in the finish formed article.

It is known to add a variety of fillers to plastics materials used in injection moulding processes. Generally they provide strength and/or hardness improvement, or merely provide economy by bulking the relatively more expensive plastics materials.

However use of fillers in the plastics materials for the present invention provides unexpected advantages.

Accordingly l prefer to add fillers to the plastics material mixture used in my present invention and indeed My Original Method.

Use of talc in particular provides advantages in that it causes the formation of complete voids in the foamed plastics material. This is believed to be caused by talc comprising calcium caiboiiate platelets. These have relatively sharp edges, which sever the bubble walls formed internaliy of the foaming plastics material. The result is the formation of a complete void, with a rough internal structure. In turn this causes greater expansion in that the removal of the internal foam structure allows greater expansion of the thick wall portions of the finish formed article. Greater expansion provides greater transverse dimension and hence greater stiffness.

Further I have noted that the fillers tend to migrate to the outer surface ofthe thick wall portions. Again this increases stiffness.

Furthermore, I have noted that use of fillers increases the time over which foaming occurs This results in greater dimensional expansion. which in turn enhances stiffness. I believe this eflect to be caused by the plastics material incorporating fillers having a greater thernial capacity. Tliis slows cooling and hence allows foaming over a longer period.

I believe the following fillers to be advantageous: Talc Chalk Alumina Glass spheres Glass and carbon fibres.

To help understanding of the invention, specific embodiments thereof will now be described by way of example and with reference to Figures 8 to 12 of the accompanying drawings, in which: Figure 8 shows a respective moulded - 8(i) - and foamed - 8(ii) - section for a cutlery handle; Figure 9 shows similar sections for another cutlery; Figure 10 shows similar sections for an arm of a garment hanger; Figure 11 shows similar sections for a reinforcing rib for a vehicle door pocket; Figure 12 shows similar sections for a reinforcing ridge for a vehicle skin panel; Figure 13 shows a =arnlent hanger manufactured in accordance with my in present invelition, with scrap details showing cross-sectional shape, the hanger having different left and right arms; Figure 14 shows in cross-section an eye bath in accordance with the invention; Figure 15 shows a side view ofa knife manufactured in accordance with my present invention; Figures I Sa and 15b are respective cross-sectional views on the lines A and B in Figure 15 of the knife showing foamed sections; Figures l6a and l6b are similar views of the "as-moulded" sections prior to foaming expansion; Figure 17 shows a side view of a fork manufactured in accordance with my present invention; Figure 18 shows a plan view owt the fork of Figure 1 7; Figures 1 ()a, I ')h and 1 ')c are r espective cross-sectional views on the lines A, B and C in Figure 1 8 of the fork showing foamed sections, Figures 20a, 20b and 20e are similar views of the "as-moulded" sections prior to foaming expansion; Figure 21 shows a plan view ofa spoon manufactured in accordance with my present invention; Figures 22a and 22b are respective cross-sectional views on the lines A and B in Figure 2 1 of the spoon showing foamed sections; Figures 23a and 23b are similar views of the "as-moulded" sections prior to foaming expansion;

Figure 24 shows a plan view of another spoon manufactured in accordance with my present invention; Figure 25 is a plan view of the spoon of Figure 24; Figures 26a and 26b are respective cross-sectional views on the lines A and B in Figure 25 of the spoon showing foamed sections; Figures 27a and 27b are similar views ofthe "as-moulded" sections prior to foaming expansion; Figure 31 is a side view of a fork as moulded, i.e. of the mould cavity shape, with a series of views showing cross-sectional shape; Figure 32 is a siniilar side view ofthe finish formed fork, with the same series of cross-sectional views showing the shape after foaming expansion and an enlarged scrap view of one cross-section; Figure 33 is a plan view of the fork; Figure 34 is a cross-sectional plan view of a mould for a handle; Figure 35 is a cross-sectional view on the line XXXV-XXXV in Figure 34 for the mould; and Figure 36 is plan view of the handle, Figure 37 is a side view of the handle; Figure 3S is a cross-sectionai view of the handle at the position corresponding to the line XXXVIII-XXXVIII, showing a circular cross-section; Figure 39 is view similar to Figure 37 of a modified handle being formed to flattened shape; Figure 40 is another similar view of an elliptical handle shape; and Figures 41,42 and 43 are views similar to Figures 17, 19 and 20 of another fork; and Figure 44 is a cross-sectionai view of a pre-fonn and a corresponding expanded section of a knife blade.

It should be noted that much of the following description is merely of cross- sectional shapes of the mould region(s) whose mould part gap is less than the thickness of the said portion(s) of the finish formed article and the corresponding shapes of the finish formed articles. Moulding of the articles and their foaming expansion can be

carried out in accordance with My Original Method, as exemplified with reference to Figures 2,3, & 4.

Turning first to Figure 8, the mould shape 8(i) in which the cutlery handle skins has double tear drop 381,382 shape, with a narrow, intervening thin wall portion 383 at the thin ends 384 of the tear drops. These taper 385 out to domed ends 386. The foamed shape 8(ii) retains the intervening thin wall portion 383. However the tapered portions 385 have blown outwards, adopting a substantially constant radius of curvature 387. The domed ends 386 have had their radius of curvature reduced to that ofthe blown tapers. The overall height 388(i), 388(ii) ofthe cross-section has fore- shortened on foaming expansion, whilst the width 389(i), 389(ii) has increased. At the junction of the foamed skins 390 and the thin wall portion 383, the opposite skins have been pulled pulled apalt and bent so as to extend locally away from each other at the limit of the extension of the still liquid central portion of the plastics material on mould opening. The point at which this occurs in the finished article can be controlled by the degree of cooling of the thin wall portion and the skins, and the timing of mould opening.

Figure ') shows a mould shape 9(i) based on the letter W, with a thin wall portion 392 joining two half round thick wall portions 393',393" Each of these has a mould pait gap approximately double that of the thin wall portion. The corners 394 at the ends of the concave sides 395 of the half round portions are radiused. On foaming expansion, the concave side skins 396 pop over-centre to become convex. Whilst symmetry of the final shape with respect to the medial plane 397 of the thin wall portion, is more definitely assured if the skin length 398 from the thin wall portion via the initially convex skin 399 to the remote corner 400 is the same as that via the concave skin; this is not necessary. The convex skin length 399 is likely to be longer, with the result that the point 400 is drawn out of the medial plane. The skin can be expected to stretch to a limited extent only, so that the radius of curvature 401 of the finish formed handle will be approximately that of the original convex skin.

Figure 10 shows a mould shape 10(i), which is based on the letter X. It has no thin wall portion. Each of the four arms 402 of the X has a smooth, tightly radiused

end 403. Their roots 404 merge with a larger radius curvature 405, providing a central reservoir 406 of liquid plastics material able to foam after skinning. It will be appreciated that this mould shape provides a relatively large overall circumferential length, so a relatively high degree ofvolumetric expansion is possible to the foamed shape 10(ii). This is advantageous in providing a rigid arm for a product such as a garment hanger.

It should be noted that all three of the embodiments just described can be incorporated in elongate articles of a length which can be chosen at will. In the case of the cutlery handles, thin wall portions will be incorporated at the end of the handle to provided a blade, for instance ib a knife The garment hanger, on the other hand, may not have a thin wall section. The cross-sectional diameter and the curvature of the central axis of the elongate section can be chosen to suit the garment. It is anticipated that the section can be branched to provide the hook of the hanger.

Figure 11 shows a mould shape 11 (i) for a reinforcing rib for a vehicle door pocket. The shape is based on the letter Y, with tightly radiused ends 411 and concave sections 412 to provide for appreciable expansion on foaming. The foamed shape l 1(it) is symmetrical with respect to a thin wall portion 413, providing the wall ofthe pocket. If this is to be biased to one side, for instance the inside ofthe pocket to provide enhanced grippability the foaming expansion can be arranged to occur whilst the thin wall is still in contact with the outside shape defining part of the mould. For this, the Y 11 (iii) may be biased inwards so that the mould face 414 can be arranged to provide a flat surface for the foaioied rib I I (iv) to fonn against.

Figure 12 shows a mould shape 12(i) for forming a rib 421 on the inside of a thin wall car body panel 422. The mould shape is generally M shaped, with a central concavity 423. This passes over-centre on foaming expansion 12(ii), whereby the rib balloons away from the panel. It should be noted that the rib has a semi-circular shape, which is nevertheless of substantially constant radius of curvature.

As mentioned above, the work which 1 have carried out on My Original Method has been with polypropylene, using an endothermic blowing agent releasing

carbon dioxide gas. Polypropylene is a material which is amorphous above 140"C, its crystallisation temperature. I prefer to inject it at 2200C, so that it flows very readily into the mould and is free to expand in the mould cavity to the extent that this is under- packed. Skinning involves cooling of the material to below the temperature at which crystallisation can occur. I believe that as soon as the skin is urged to stretch by the foaming expansion after mould part withdrawal, the skin is liable to crystallise. Since the pressure which can be exerted by carbon dioxide is limited (it liquefies at 2.24 bar at room temperature), the crystallisation resists substantial stretching of the skin. Even if crystallisation is not the mechanism inhibiting skin stretching, the solidification of the skin does this.

The invention Is not restricted to polypropylene. With other materials, which are less pioiie to c -vstallisation, nlol-e skin stretch may be experienced. Nevertheless, this does not affect the desirability of forming the original skin with excess circumferential length and expanding this to at least part circular form.

Referring to Figure 13, tlie hanger there shown has a hook 450, a right arm 451 and a left art1 459. These linibs are all expanded In accordance with the invention to include portions expanded to circular cross-section and having central voids. The hook and right arm have iteiierally C-shaped pre-form shapes 4501,45 11, which expand to circular shapes 4502,45 12. The left arm has a flatter pre-form cross-section 4521, with a central thin land 453. On either side ofthe land the pre-form has thicker wings 454, which thicken outwards, with a circular contour 455 on the same side as the land 453. The opposite side 456 is coinplementarily concavely shaped, with a curvature such that it can expand ollt to continlle the convex curvature 455. On expansion, tlie bi-lobe shape 457 fornis, with two circular edges 458 and the central land 453 tangential to these.

Referring to Figure 14, the article there shown is an eye bath 470, which is elliptical in plan - not shown. The bath has a thin wall 471 and a hollow annular (subject to the eiliptical shape) riin 472. The latter is shown at one side of the Figure, whilst the tear drop pre-form shape 473 is shown at the other side. This expands to hollow circular cross-section with a circumferential length 474 equal to that of the tear

drop 475. This results in a slight fore-shortening 476 of the height of the eye bath from the height of the pre-form. The eye bath is of flexible polypropylene material, where the rim can deform to the eye socket and seal in eye wash solution in use.

Referring to Figure 15, the knife 500 there shown has a blade 501 with a thin wall/gauge thickness 502 tapering down to a cutting edge 503, see Figure 15a. The back of the blade has a thick foamed edge 504 of substantially circular cross-section.

This starts at the distal end of the blade and continues into the handle 505, round the distal end 506 of the handle and terminates at a point 507 where the handle has widened to blend with the blade. The handle has a thin web 508 extending along its full length.

It will be appreciated that the knife has high stiffness in bending, not only in the normal cutting plane, due to the I beam configuration of the handle, see Figure 1 5b. In the perpendicular plane, the foamed back 504 of the blade and the foaming expansion 509 to the point 507 provide substantial stiffness.

The foaming expansion along the handle and at its distal end provide a comfortable feel to the knife Turning to Figures loa and l6b, the cross-sectional shape of the mould cavity for producing the knife is showii. It will be noted that at the edge of the foamed part, the thick section has a peripheral surface 510 which is perpendicular to the central plane 511 of the knife. This plane is the joint plane of the mould for the tool, which is symmetrical about the joint plane. The thick section tapers 5 12 to the central thin wall ofthe knife handle and of the knife blade, to provide an adequate surface skin to inflate on foaming expansion to substantially circular cross-section. It should be noted that the cross-sectional height 5 13 of the handle is greater as moulded than after foaming expansion 5 14, due to fore-shortening of the thick portions on foaming expansion.

Figures 17 and 18 show a fork 520. It is of largely conventional shape, with prongs 521, a swan neck 522 and a handle 523. The prongs are foamed. The outer prongs 521' and the edges 524 ofthe swan neck and the edges 525 of the handle have

a continuous foamed rib 526 extending to and terminating at the distal end 527 of the handle. The foaming expansion of the inner two prongs 521 " terminates at their root in a thin wall palm section 528 of the fork. At the swan neck 522, a central foamed rib 529 provides added stiffness. This rib extends into the handle and divides into two subsidiary central ribs 529'. The handle thus has three thin wall webs 530. To give the handle a comfortable feel, it is curved in transverse cross-section.

Figures 19a, l9b, 19c, 20a, 20b, 20c show typical foamed and corresponding as moulded - prior to foaming - cross-sections.

Figure 21 shows a spoon 540. It has foamed edges 541 extending from its bowl 542, through its neck 543, and along its handle 544. A central foamed rib 545 extends from the bowl along the handle, to give added stiffness, particularly to the neck.

Figures 22a, 22b, 23a, 23b, show typical foamed and corresponding as moulded - prior to foaming - cross-sections.

Figures 24 and 25 show aiiothei spoon 550 This has a modified invert channel section handle 551, with lateral flanges .559 which are of thin wall thickness, and a foamed central web 553 Figures 26a, 26b, 27a, 27b, show typical foamed and corresponding as moulded - prior to foaming - cioss-sections.

Refen-ing to Figure 31, it should be noted that not only is the shape shown that of the mould cavity, it is also the shape of a fork moulded in the cavity without foam producing additive; whereas the shape of the fork shown in Figure 32 is one from the same mould cavity with foam producing additive in the plastics material.

ln general terms, the fork has a set of prongs 1001, a swan neck 1002 and a handle 1003. The prongs 100 1 have extensions 1004 into the swan neck, which has

web areas 1005 between the extensions. The extensions 104" 1043 of the inner two prongs 1012,1013 extend back to the forward end of the handle. Similarly, the extensions 1041,1044 of the outer two prongs 1011, 1014 extend back peripherally of the swan neck to the handle. In addition, the two outer prongs 1011,1014 have extra extensions 1411,1441, which branch from the prongs close to their outer ends and extend back into the swan necle between the peripheral extensions 1041,1044 and the inner prongs' extensions 1042,1043. The extra extensions taper out before reaching the handle.

As apparent from the cross-sectionai views in Figure 31, the prongs taper to pointed ends 1 006. The inner prongs 1012, 1013 and their extensions 1042,1043 are of constant, eguilateral trialnle cross-section behind the tapers 1007 of the ends 1006.

The web areas on either side of tlie extensions 1042,1043 are an order of magnitude thinner than the cross-sectional dinieiisiois of the extensions, and so can be ignored as regards effect on the foamed shape. Tlie extra extensions 1411,1441 also are of equilateral cross-section. The outer prongs 1011,1014 and their extensions 1041,1044 are of right angle triangle shape of height - from the web areas - equal to the inner prongs. The handle shape is more complex. Essentially it is square edged 1081, with narrow top and bottom retllrns 1082, 1083 extending in from the edges at right angles.

The main extent 1084 ofthe top is symmetiically ramped up to a central rounded top ridge 1085. The bottom is of coniplementary shape 1085. except that it has a central fat 1086. The top is thais convex and the bottom is concave.

Referring now to Figure 32, the foamed shape of the fork is as follows: The very ends of I 006 of the prongs are of solid plastics material. Where the tapers 1007 reach a certain tliickiiess, foaming is present in their centre, although without any apparent change in shape frown the Figure 31 shape. However where the full section is reached, surfaces of the prongs and their extensions bow out, as the internal foam pressure inflates them to maximize the internal volume of the cross- section. However the initial skin shape gives little scope for marked increase in the internal volume. The result is that almost the entire extent of the prongs and their extension is of cellular phase. The first exceptions are the points 1006. The second exceptions are the regions of intersection of the outer prongs 1011,1014 with their

extra extensions 14 II, 144 1. At the intersections, regions 1040 of sharp convexity are formed in the original skin, which enables a substantial expansion in the internal volume ofthe moulding at the regions. Figure 32 shows that this results in a substantially circular cross-section 1401 with a central void or hollow 1402 developing. The original skin shape was complex, with sharp features. These are not entirely smoothed out, so that the cross-section has discontinuities in the circular shape.

The handle also has a concavity in its original skin shape, in the form of the shape 1085. Thus this can invert to substantially circular cross-section on foaming expansion and a central hollow 1800 develops. The width of the handle is much greater than the width of the outer prongs at the junction with their extra extensions.

Thus the handle expands to a much bigger diameter. Whilst the hollow 1402 has a void of the same order of cross-sectional dimension as the skin thickness; the central void in the handle is an order of magiiitude larger than the skin thickness.

As shown in the detail to cross-section vi-vi, the right angle features in the initial skin shape between the edges 108 I and tlie returns 1082, 1083 remain in final cross-section, but the edges and returns become local concave features as a result of the overall inflation of tlie cross-section Measuieinents have indicated a 3% stretch only from the skin of the handle as moulded to its Figure 2 state, when using polypropylene.

Referring now to Figures 34 & 35, the mould there shown is for a handle for an implement whose operative portion is not shown. The mould cavity has a large diameter distal end 200 1 and a small diameter proximal end 2002 at which the mould would be extended for the non-shown operative portion of the implement. Between the ends, the cavity has straigllt sides 2003, which approach each other. The mould has a uniform gap 2004, corresponding to the wall thickness of a handle to be moulded if plastics material without foam producing additive is used. Small central and peripheral grooves 2005,2006 are cut into the two parts 2007,2008 of the mould.

The handle 20 10 moulded with foam producing additive is shown in Figures 36, 37 & 38. On injection, the skin 2011 is formed at the internal surface ofthe cooled mould, which imparts the initial skin shape. On opening of the mould prior to complete cooling of the plastics material at the centre of the moulding, the material expands by foaming to the shape shown. It should be noted that the skin does not appreciably extend in length or surface area; rather, it is inflated to shape. The skin is smooth on the outside, with no foam bubbles forming at the surface, because this solidifies whilst the moulding pressure is maintained and inhibiting foaming. However within the skin, bubbles or cells 20 i 2 are present. These increase in size until the central void 2013 is reached. Throughout most of its length, the handle has a circular cross-section as shown ill Figure 38. The outer surface is not truly circular, having the shape of the grooves 2005, 2006 - albeit somewhat deformed to the circular shape.

Where the cross-sectioii is circular, the sides 2003 are straight; although it should be noted that they are closer together than originally. At the large diameter end 2001, the curved shape into which the skin is formed initially inhibits the full circular cross- section from developing. Thus the thickness 2014 ofthe handle at its centreline progressively decreases close to the end. The result is that the sides 2003 curve gently out 201 5 to meet the curvature of the end. The thickness 2014 begins to decrease from the same position as the sides curve out. Similar etfects occur at the small diameter end 2002, biit they are not so prolililient Turning now to Figure 39, the elt'ect on the cross-sectional shape at the longitudinal centre ofthe handle can be seen, when the mould parts 2007,2008 are opened to a limited extent 2020 only until the handle has set. It expands against the mould parts, which impress their central shape on the handle, that is to say the flat surface of the mould part 2008 and the similar surface with the central groove 2005 in the case of the other mould part 2007. This has the advantage of spreading the edges 2003, with consequentiai increase in monient of inertia; but the disadvantage of increasing cycle time due to the time taken in holding the mould parts partially withdrawn for the expansion to occur. A similar effect can be achieved with expansion in free air by reducing the amount of foam producing additive in the plastics material mixture, as shown in Figure 40. In this case the cross-sectional shape does not

develop to a full circular shape, but approximates to an elliptical shape 2030. This shape can be produced with the same cycle time as with the shape of Figures 35, 36 & 37.

Turning now to Figures 41, 42 and 43, there is shown another fork 2101 with a pronounced curvature of its swan neck 2102, with a local radius of curvature 2103. A characteristic of the invention particularly where voids are blown, is that the relevant portion of the article tends to straighten, as with inflation of a child's balloon. This fork incorporates edge flanges 2104 in the swan neck, which extend radially, that is transversely of a wall 2 105 extending through the swan neck. The pre-form - Figure 43a - of the swan neck has a concavity 2106 facing in from each flange and extending along a narrow web 2107 between each flange and the wall 2105. On expansion, the concavity's skin passes over centre. However, since it is facing essentially laterally and its height in radial direction does not increase, there is little tendency for the expansion to straighten the swan neck. This effect is enhanced if the fork is held on the tool whilst the expansion occurs. with the skin 2108 ofthe web and flange in contact with the tool being kept relatively rigid by cooling from the tool. The central tines 2110 of the fork also incorporate curvature limiting flanges 2111, at least at their roots. These flanges have webs 2112 and lateral concavities 2113 at each side. Further, the webs have a groove 2114 - formed by a rib in the tool - extending into the flange. This groove is kept cool during expansion; so that not only does it not invert and expand, but also the groove helps to maintain the curvature e of the tines.

Turning on to Figure 44, there is shown a pre-form and expanded cross-section of a knife blade. The pre-form is stepped 2201 in thickness from the wall thickness of the blade 2202. The back 2203 of the blade is of C-shaped cross-section. On expansion, a void 2204 develops at the back, with the skins 2205 of the thicker wall being pulled apart as far as the step 2201. With excess of blowing agent and a relatively short cooling time before the expansion is initiated, the expansion will be to circular cross-section with the step 2201 being within the circular circumference.

However, control of the blowing agent quantity and cooling time, the expanded cross- section as shown will develop, with a semicircular shape tapering in to the step 2201.

The void 2204 develops at the back and a cellular phase is present between the skins 2205, at least adjacent the step 2201.