Model Engineers Workshop No.161, March 2010.

How to Machine Castings

Harold Hall makes a Keats Vee angle plate

A reader of the magazine commented to theeditor that, as a novice metalworker, he foundit difficult to decide which face to work on firstwhen confronted with a casting to machine. Atthe time he was considering the Keats AnglePlate (photo 1) shown in The CollegeEngineering Supply catalogue.1 The purposeof his letter was to suggest that an articleexplaining where to start would be helpful tohim and no doubt many others.

Having agreed to provide the article it wouldbe good to be able to say, mount the castingin a manner so that you can first machine thesurface that ..., but of course it is not thatsimple. The very nature of a casting is that itsshape in most cases is likely to be quitedifferent from any other casting the workshopowner is likely to be called upon to machine,the Keats angle plate being a good example.

The reader having mentioned this item makesit an obvious choice to use as a subject so asto come up with some pointers for this andother castings that need to be machined.

So as to provide other examples, in the nextissue I have chosen the Small Vee Angle Plate,also from The College Engineering Supply.Both items are intended to perform similarfunctions though the Vee angle plate is moreadaptable and will work more readily withsmaller workpieces.

I therefore intend, in addition to detailing thesetups for machining these, to also explaintheir purpose and the pluses and minuses ofeach design. The castings as received areshown in Photo 2, with those for the Keats onthe left and the two for the Vee angle plate onthe right. In both cases the hardware requiredhas to be provided by the workshop owner.

1 http://www.collegeengineering.co.uk/

The nature of a castingI do not confess to being knowledgeableregarding the finer details of metal casting, butin this project we are considering just ironcastings, which are almost exclusivelyproduced in a sand mould. In very basic termsthe mould is made in two parts that can beseparated so as to enable the pattern to beremoved when, after reassembling, themolten iron can then be poured to producethe required shape.

Sk.1 shows a very simple example of arectangular section component being cast. Istate rectangular, but I think all readers willunderstand that the sides have to be taperedso as to enable the pattern to be removedwithout damaging the sand cavity prior to theiron being poured. Because of this the castinghas two parallel faces and the remaining facesare tapered and non parallel. The two parallelfaces would be relatively easy to secure in avice, depending on size of course, and othermethods such as using an angle plate wouldalso be quite easy, although the tapered faceswould present a problem. The Vee blockcasting in Photo 2 (rear right) has similaritiesto this simple example.

Castings come in a very wide range of shapesand sizes and many will not be able to bemounted so easily as the above explanation.Sk.2 shows another example and from this itwill be seen that in this case no two faces areparallel and as a result, mounting it in a vicewill be much more of a problem though notnecessarily impossible. It is because of thesevariations that makes it impossible to give anall embracing answer to the readers question.I will though explain for each of the fourcastings how I come to the decisions I do, bothwith regard to the face to machine first andthe sequence for the remaining faces. Ofcourse, many castings will not have an obviousfirst face and some workshop owners willchoose one whilst others will choose another.

1

The first taskFirst, one must study the casting carefullynoting which faces are parallel and which arenot and which faces have to be machined, asoften many are left in their as-cast state.

Having spent time considering the situationthe aim at this stage would normally be tochoose the most secure method of holding thecasting whilst presenting the first surface formachining. The choice though should also takeinto consideration whether or not the surfacemachined first will make subsequent stageseasier to mount. Therefore, where a numberof choices for the first surface would appearequally acceptable the decision must be madeon the basis of the next operation. Planningahead is an essential feature when machininga casting!

Also included in the considerations is that oneshould attempt to choose a method thatpresents the larger surfaces in a manner thatenables them to be machined equally overtheir full surface. If this is not done, then asmore is removed on one face than the otherthe increased amount of metal removed maymake it impossible to work to the dimensionsgiven. This I feel is best understood byreference to Sk. 3. In this, A shows thecasting as supplied and which needs to besurfaced on the two outer faces to make themat 90 to one another. With the short facemounted vertically, as at B, much more hasto be removed compared to the longer facebeing mounted vertically, C. Having firstmachined the casting as at C it can then bemounted as shown at B and machinedequally over the larger surface.

Fortunately, the pattern maker will havechosen in most cases to make the taper on theshortest sides, see Sk. 4A, so as to reduce thedepth of the pattern in the mould andminimising the amount of metal that will haveto be removed at the machining stage if madeas per Sk. 4B. Another benefit of this is that itreduces the amount of iron used therebyreducing the cost of the casting as a result.

With castings that conform to the above, oneof the shorter faces is very likely to be the oneto machine first. That is certainly the case withthree out of the four castings being machined

in this article. I should add here that theangles shown in the sketches are, for clarity,more than those normally present on thecasting which is usually between one and twodegrees per face.

Machining Cast IronThe subject of machining cast iron has beencovered frequently in the magazine andtherefore a very detailed explanation isprobably not appropriate. However, as thearticle is aimed at the workshop owner new tometalworking I will briefly cover the mostimportant points.

1) The surface of iron castings canfrequently posses hard spots that willrapidly blunt high speed steel tooling.

2) Because of the hard spots, if possibleuse carbide tipped tools for initiallymachining the castings surfaces.

3) If the surface starts to take on a glazedappearance this is because the depthof cut is not sufficient to get below thehard skin.

4) This can either be because thecastings surface has a shallow hollowor that it has been mounted on themachine with the surface beingmachined sloping down very slightly.

5) In this case, return to the start,deepen the cut, and recommence tomachine the surface.

6) If you do not have a tipped tool,reserve an old HSS cutter for the task,keeping your sharp cutters for moreappropriate tasks.

7) If you have to use a HSS cutter, grind achamfer, about 1 to 2mm, around theedge of the casting so that the cutterdoes not have to break through thehard surface.

8) Whilst less important, the chamfer isalso worthwhile if using a tipped tool.

9) After the hard surface has beenremoved, iron castings machine easily.A sharp HSS cutter can then beemployed to machine to the sizerequired and or to achieve a betterfinish.

10) As a result of the surfaces having beenmachined, the internal stresses withinthe casting may have changedresulting in the casting distorting. This

2

can happen between the initial stagesor in storage after it has beencompleted.

11) Machined castings are therefore bestleft for a few days (preferably more iftime permits) and then finallymachined to size, see point 9 above.

The Keats Angle Plate

The Main CastingThis conforms very closely to that shown in Sk.3 with the two faces of the larger web beingessentially parallel and the shorter webtapered on each face clearly indicating whichway the casting was produced in the mould.

The drawings provided by College Engineeringshow the casting being machined on the twoouter main surfaces plus the two faces of theVee. Additionally, two parallel slots have to befully machined as they are not cast in.However, I chose to change the centres of theparallel slots to 60mm from the 70mm shownon the drawing as they would then line upwith the slots on the Quick-Set Face andAngle Plate from Hemingway Kits that wasfeatured in MEW 155.

Surprisingly, the drawing does not call for thetop of the web opposite to the main face to bemachined but as it may be used as a locationfor the workpiece in a few cases I considerthat it should be. The ends of the shorter webare also tapered in view of the way the itemwas cast but are not shown as machined onthe drawings. However, just possibly, the endsmay be useful as reference faces or a means oflocating it if stood on its side so I decided tomachine these also. It would also provide achallenge getting the Vee central to the twoends that would add to the depth of thisarticle.

Having considered the points that I havealready discussed it was obvious that thecasting should be located off its larger facewhilst the shorter one is machined. However,it was first necessary to remove any localisedhigh points (pin head size or a little bigger) andbeing very few I removed these using a file.With that done the face was checked against a

surface plate and found to be adequately flatfor the next task.

Should you find it necessary, any high pointsshould be removed either with an old file orby localised grinding. If the latter do take carenot to remove so much that you are below thelevel to which you will ultimately wish tomachine it.

Perfection will not of course be possible so apiece of thin hard card between the castingand the mounting surface will compensate forminor errors. However, do not be fooled intothinking that the situation is better than it is asoften the casting will fall away towards itsedge. Because of this do ensure that any barclamps are applying their force well in fromthe perimeter of the casting.

With those precautions having been takennote of, I machined the two faces as shown inPhoto 3 , having set them reasonably level sothat they were machined equally. Regularreaders of my articles may be aware of mycomments that relying on a single clamp is tobe avoided if possible and that one too manyis much better than one too few. In this casethough, the compact nature of the assemblymade it acceptable. However, do note that themachining forces are towards the angle plate,not away or across. I did consider that someworkshop owners would not have an angleplate large enough to permit more clamps tobe employed so decided that I would like toillustrate a method within the bounds of whatwould likely be available in some cases.

It may seem obvious that the next face shouldbe the large main face, but I decided that thetop of the web should now be tackled. Thereason is that it was very likely that the mainface and the top of the web would only benominally parallel, so by machining the webfirst the two surfaces were trued up with theminimum of metal removed (Photo 4).

Next, I chose to mount the casting onto twosquare posts when machining the main face(Photo 5). To do this the surface machined inthe last operation was placed firmly on themachine table ensuring that the larger facenow being machined was parallel with the topof the web already machined. It is worth

3

considering that posts such as these can findmany uses on the milling machine, so areworth making.

You will see that I have used a form of clampthat I rarely use but their pivoted end piecemakes them ideal for dealing with the slightlysloping surface of the casting. Again note thatthe cutting forces are towards the posts.Whilst the casting is set up in this way it wouldappear an obvious time to make the two slotsmentioned earlier, but as they need to beequally spaced about the Vee, not yetmachined, and being super cautious, I delayedthe process just in case the Vee did not end upin the position I anticipated.

The next task was to machine the two faces ofthe Vee being, I thought, relatively easy, otherthan to measure the result during themachining operation. Unfortunately though, Ihad overlooked the fact that the projection ofthe cutter from the milling spindle wasinsufficient to avoid the top edge of thecasting fouling the underside of the millingspindle. I could have used an old high-speed-steel end mill in the cutter chuck, which wouldhave given me the necessary clearance andwould anticipate that this would be themethod chosen by most.

Having already machined the surfaces ateither end of the Vee about to be made, thecutter would not be breaking through a hardskin so I did not anticipate too much of aproblem with this method.

In an endeavour to stay with tungsten cutters Idecided, as an experiment, to try using myboring head for the purpose and even thoughthe cut was intermittent, the arrangementworked quite well providing the depth of cutwas kept to a reasonable value; 0.3mm I seemto remember.

I marked the two machined faces withmarking blue and scribed machine-to lines todefine the width of the Vee. This being adistance of 60mm, I positioned it to be anequal distance from each end of the casting,nominally 30mm. I mounted the casting onthe angle plate setting the faces to 45, doingthis away from the milling machine on thesurface plate as the larger surface makes it

easier to use the protractor from mycombination square. Having done that it wastransferred to the milling machine table andthe Vee made using the boring head asmentioned above (Photo 6). When machiningthese faces one can stop just short ofcontacting the other as a groove is machinedin the bottom of the Vee that will remove anyunmachined surface.

The casting was next tilted the other way andthe second surface machined. In this case themachining forces were away from the angleplate which is not ideal but this was difficult toavoid and with the light cut being taken therewas no problem. With the angle plate still onthe machine table the casting was moved tothe horizontal position and the groove in thebottom of the Vee made.

Having finished the Vee I was now able tomachine the two fixing slots ensuring thatthey were equally spaced about it. To do this Iclamped a piece of round material in the Veewhich gave me a reference point for settingthe position of the slots. Again using thesquare posts, Photo 7 shows the set up withthe first slot already machined.

This now leaves the ends of the web. Theyneed a rather special setup to ensure that theVee is central to a fair degree of accuracy. Thecasting was positioned on a round post fixedto an angle plate and set vertically as shown inPhoto 8 and then machined. Next I rotated thecasting through 180, setting it vertically oncemore and machined the second end using thesame down-feed setting as for the first (Photo9). These operations could not though becarried out at this stage as they needed theclamp and its fixings to be completed.

The ClampThis is a casting with no obvious bestsequence for machining. I am sure somewould tackle it differently. The parallel facesare those of the raised bosses with the sidesof the Vee being tapered and narrowest at thepoint. Having removed any major roughnesson the sides I chose to hold this in the vice butusing some copper shims to help compensatefor the taper and any localised raised portions.With the vice just lightly closed I used a softhammer to encourage the casting to take up a

4

horizontal position, checking this using theprotractor from my combination set.Incidentally, the clamp is a malleable ironcasting and machines quite differently fromthe main casting.

As my milling machine has not been set upaccurately to have a level table, I first placedthe protractor on the table and adjusted it toread zero. With that done I then used it toposition the casting thereby compensating forany error with the machine table, of courseprecision is not a requirement. The vice wasthen fully tightened and in view of the limitedamount of machining necessary I was happy tomachine the two bosses using this setup(Photo 10).

It is worth noting that security can in manycases be increased by a shim of soft copper,about 1mm thick, that will, under pressure,conform to the minor irregularities of thecasting, improving the grip appreciably. This isespecially beneficial if working as in thisexample.

I then turned the casting over and againmounted it in the vice, but this time used aparallel to support it, ensuring the two endswould machine to the same thickness. Whenthe Keats angle plate is in use for smalldiameter workpieces, the clamp is invertedwith pressure being applied by the top of theVee. To avoid this being a sharp edge, a 3mmwide flat was made which was also machinedwhilst the casting was still held in the vice.

Next I mounted the casting onto the squareposts, seen previously, and machined the firstside. The casting was then turned over andthis time making sure the already machinedside was in contact with the machine table,the second side also machined (Photo 11).This ensured that the two sides were parallel.

The two sides of the Vee and the groove in itsbase were the next to be machined using asimilar method to that for the main casting.However, the absence of a flange made it alittle more difficult to clamp securely. Photo 12shows the setup I used. In this I have againused one of the posts which I set accurately at45 on the angle plate before being mountedon the machine table. The part was then first

clamped to the post thereby easily setting it atthe required angle with the bar clamp finallybeing added.

From this photograph it can be seen just howclose the bottom of the milling machine quillwas to the ends of the casting and why thecutter would have been inadequate formachining the Vee as seen in Photo 6.

I have not mentioned the tapped holes in themain casting or the plain holes in the clampbut produced these at this late stage. Someworkshop owners would I am sure have drilledthen earlier and used them to fix the castingsto the angle plate rather than using themethods I have illustrated. If you have a largeenough angle plate then that is certainly amethod worth considering providing the slotsin the plate are suitably spaced.

The clamp now being finished I was able tomachine the ends of the main casting as wasexplained earlier and illustrated in Photos 8and 9. You may though understandablyconsider this an unnecessary feature andchoose to bypass this operation.

Having done that, there just remained for theedges to be lightly chamfered and the non-machined surfaces painted. With that, wehave the finished item (Photo 1), that is unlessI eventually choose to re-machine some of thefaces as explained below.

At the time of writing this article, I have notchecked to see if there has been anymovement within the casting as a result of themachining, as this is best undertaken after atime delay of at least a few weeks. The onlyarea that is important is that the two surfacesof the Vee are at 90 to the main face, asituation that I will check using an engineerssquare. If I find an error in access of what I feelis reasonable I will set up a cylindrical squareon the milling machine table and clamp theKeats angle plate to this using its own clampand with that done very lightly re-surface justthe main face.

In the next issue I will describe how tomachine the Vee angle plate castings.

5

Figure 1. The finished Keats angle plate.

Figure 2. The castings as supplied, both for the Keats angle plate and the Vee angle plate.

Figure 3. The first faces to be machined are those on either side of the Vee.

Figure 4. Machining the second face.

Figure 5. The main face was machined with that in Photo 3 against the square posts and that in Photo 4 firmly against themachine table.

Figure 6. The cutter used in the previous photographs had insufficient reach to machine the Vee, so my boring head wasused instead.

Figure 7. Using a slot drill to produce the fixing slots. The round bar gives a reference point for positioning them relativeto the Vee.

Figure 8. Setting the casting vertical for machining the ends.

Figure 9. The second end being machined having been set up in a similar manner to that in Photo 8.

Figure 10. Facing the fixing lugs on the clamp.

Figure 11. The square posts again being used whilst the sides of the clamp are being machined.

Figure 12. The Vee being cut.

The nature of a castingThe first taskMachining Cast IronThe Keats Angle PlateThe Main CastingThe Clamp