Checking the finished oval

An oval can be the right length, the right width and the right girth and still be the wrong shape. Its end arcs might be rolled a little tight and its side arcs a little flat, or the reverse, and the errors cancel in the three easy measurements while the outline is visibly out. A lid made to the drawing will not fit it, its rim rod will not lie evenly, and its capacity will be off by a few litres. Checking an oval therefore needs more than a tape across its axes: it needs the oval’s own shape, tested at the places where the shape is defined.

This chapter sets out the checks for the running case’s finished trough: templates at the rim and the bottom, the girth by tape, the axes and the tangent points, the flare and the level of the rim, the capacity and the leak test. It brings the earlier chapters’ checks into one inspection sheet, and ends with a trough whose side piece was rolled flat and which passed a check made only across its axes.

Templates at the rim and the bottom

Chapter 2’s templates carry the oval’s two radii. On the finished trough they are used at two levels, because the radii change from rim to bottom.

This is an illustration, with the shop’s templates and their written rises:

• rim, end: an arc of 300 mm radius on a 400 mm chord, rising 76.4 mm;
• rim, side: an arc of 1,575 mm radius on a 1,000 mm chord, rising 81.5 mm;
• bottom, end: 225 mm radius on a 300 mm chord, rising 57.3 mm;
• bottom, side: 1,500 mm radius on a 1,000 mm chord, rising 85.8 mm.

Each template is held against the inside of the wall, just below the rim or just above the bottom joint, square to the wall, at the middle of each arc. A gap at the template’s middle means that part of the wall is flatter than the drawing; a gap at its ends, that it is tighter. The shop’s illustrative acceptance is a gap of no more than 2 mm on any template.

The templates are made to the mean line, so on the inside of 1.6 mm sheet a perfect wall shows a gap of a fraction of a millimetre, which the acceptance allows for. A shop that wants a tighter check cuts its templates to the inside radii, 0.8 mm smaller.

Figure 1Templates for the finished trough

Template	Radius	Chord	Rise	Where it is used
Rim, end	300 mm	400 mm	76.4 mm	middle of each end arc, just below the rim
Rim, side	1,575 mm	1,000 mm	81.5 mm	middle of each side, just below the rim
Bottom, end	225 mm	300 mm	57.3 mm	middle of each end arc, just above the bottom joint
Bottom, side	1,500 mm	1,000 mm	85.8 mm	middle of each side, just above the bottom joint

Checking the pieces before assembly

The cheapest place to find a wrongly rolled piece is before it is riveted. Each of the four wall pieces is checked as it comes off the rolls, standing on its bottom edge on the bench, against the two templates for its arc: the rim template just below the top edge and the bottom template just above the bottom edge. Because each piece is part of a cone, its curvature changes from top to bottom, and both templates must fit for the piece to be right.

This is an illustration. An end piece is checked with the 300 mm and 225 mm templates, a side piece with the 1,575 mm and 1,500 mm templates. A piece that fits the rim template but not the bottom one has been rolled as a cylinder, not as a cone: the rolls were set parallel instead of skewed, and the piece’s bottom edge is too open. A piece that fits neither has the wrong roll setting altogether.

The pieces are also checked for their edge lengths before forming, along the rim and bottom edges of each flat pattern, against Chapter 6’s figures. A pattern that is right flat and wrong rolled points to the rolls; a pattern that is wrong flat points to the marking out. Separating the two saves time hunting for the cause.

On a batch, the first piece of each kind is checked with both templates and the rest with the rim template alone, unless the rolls are reset. Any reset of the rolls, for this job or another, means the next piece is checked as a first piece again. That rule is the one the error later in this chapter shows the need for.

Girth, axes and tangent points

Three measurements across the trough confirm its overall size.

1. Girth: a tape round the outside of the wall just below the rim should read the outside girth, 4,383.7 + 2π × 0.8 = 4,388.8 mm, as Chapter 3 found. The tape is pulled snug round the ends, where it would otherwise bridge the tight curve.
2. Axes: inside the rim, the length across the long axis should be 1,800 − 1.6 = 1,798.4 mm and the width across the short axis 898.4 mm, the mean-line dimensions less one thickness.
3. Tangent points: the four seams lie on the tangent points, so the seams themselves can be measured. Across the trough’s width, from one side’s two seams to the other’s, the inside distance at the rim should be 2 × 264.7 − 1.6 = about 527.8 mm at each end, and along its length, from one end’s seams to the other’s, 2 × 741.2 − 1.6 = about 1,480.8 mm on each side.

The tangent-point measurements are the ones most often skipped, and the ones that catch an oval whose ends and sides have traded curvature, because they test where the arcs meet, not just the overall size.

Flare and level

The walls’ flare is checked from the bottom. With the trough level on a flat floor, a square standing on the bottom plate against the wall shows how far the wall has leaned outward at each height: one sixth of the height, so 66.7 mm at the fill line and 75 mm at the rim, everywhere round the oval. Chapter 5’s single bevel gauge at 9.46° makes the same check in one step.

The rim’s level is checked by its height above the floor at the four seams and at the middle of each side and end, eight points in all. Each should be 450 mm on the mean line, plus the rim’s own height. Chapter 6 showed what a straight-cut side pattern did to the rim; a level rim is the most visible sign that every pattern was right.

Figure 2Inspection sheet for a finished oval trough

• Do: Templates at rim and bottom, ends and sides, within 2 mm76.4, 81.5, 57.3, 85.8 mm rises
• Do: Outside girth below the rim4,388.8 mm
• Do: Inside axes at the rim1,798.4 × 898.4 mm
• Do: Seam-to-seam distances at the rimabout 527.8 mm across, 1,480.8 mm along
• Do: Flare at the rim and the fill line75 and 66.7 mm from the bottom's outline
• Do: Rim level at eight points450 mm on the mean line
• Do: Capacity to the fill line, metered or weighed439.7 L
• Do: 24-hour leak test at the brimno weeps

Capacity and leak test

Chapter 4’s capacity is confirmed by filling the levelled trough to its fill line through a water meter, or on a platform scale: 439.7 litres, or 439.7 kg of water. A shortfall of more than a few litres points to walls that have not flared fully, a bottom plate set high, or a fill line marked low, and each of those can be measured from the checks above.

The trough is then filled to the brim, 502.6 litres, and left for 24 hours, as Chapter 7 set out. Every joint is inspected: the four seams, the bottom joint all round and especially at its notched end flanges, and every rivet. The leak test also shows whether the bottom sags on the floor or stand, which is a matter for the trough’s support, not its fabrication.

Both tests are done on the first trough of every batch, and the leak test on every trough, because a single weeping rivet makes a stock trough useless in a dry paddock.

Checking the lid

A trough sold with a lid, or with a matching lid made elsewhere, has one more check: the lid against the rim. This is an illustration. The lid is made to the outside of the rim rod, with its own four-centre outline struck from the same centres, so its end and side radii are the rim’s mean-line radii plus half the sheet thickness, the rod’s diameter and a clearance. Laid on the trough, it should seat evenly all round, with its skirt clear of the rim rod by roughly the same gap at the ends and the sides.

A lid that touches at the ends and stands clear along the sides, or the reverse, shows the same kind of shape fault as a template gap, and a lid gauge is often how a customer finds one. A shop that makes both lid and trough checks one lid against the first trough of each batch, and keeps a master lid, never sold, as a gauge for the batch. A master lid tests the whole rim outline at once, which no single template can do.

Records

The shop records each trough’s checks on its inspection sheet, with the trough’s batch number, the readings, the checker’s initials and the date. For a batch, the first trough’s sheet is complete; later troughs’ sheets record the leak test, the rim level and the girth, which together catch most faults in forming and assembly. A reading that drifts from trough to trough, a girth creeping up by a millimetre each time, for instance, is a sign that a template is wearing or a roll setting is slipping, and it is investigated before it becomes a rejection.

Why the order of checks matters

The checks are cheap in a particular order. Templates take a minute and need nothing but the trough; the girth tape and the axes take a few more; the level and the flare need a flat floor and a square; the capacity and leak tests need water, a meter or a scale, and a day. Each later check is more expensive and finds faults later, when they cost more to fix.

The shop’s inspection sheet therefore runs from shape to size to function: templates first, then tape, then level and flare, then water. A fault found by a template is fixed before any water is wasted; a fault found by the leak test means draining, drying and resealing. The sheet’s order is the shop’s way of finding each fault at the cheapest moment it can be found.

The error: a side rolled flat, checked across the axes

This is an illustration. On the first trough of a later batch, one side piece was rolled on rolls that had been reset for another job and not returned to the trough’s setting. It came out with a radius of about 2,000 mm instead of 1,575. The assembler pulled it into place at the seams and riveted it.

The trough was checked across its axes, 1,798 × 898 mm, within a millimetre of drawing, and its girth was within 2 mm, because the side piece’s length was correct and only its curvature was wrong. It passed. The fault surfaced when the distributor of a matching lid found that the lid would not seat along one side: the side bowed in where it should have curved out.

A side template on that side would have found it at once. On a 1,000 mm chord, a 2,000 mm radius rises only 63.5 mm against the template’s 81.5 mm, so the template stands 18.0 mm off the wall at its middle, far outside the 2 mm acceptance. The side was cut out and replaced, and the shop’s inspection sheet was changed so that the templates come first, before any tape measure, on every first-off trough. Length, width and girth are necessary checks of an oval, and they are not sufficient: only the templates test its shape.

Figure 3Rise of a 1,000 mm chord on the side arc, by rolled radius

Drawing, R1,57581.5 mm

Rolled flat, R2,00063.5 mm

Rolled tight, R1,40092.3 mm

Bringing the oval together

Every check in this chapter traces back to one idea from Chapter 1: a four-centre oval is four arcs from four centres, and every outline in the trough, inside, outside, rim and bottom, is the same four centres with different radii. The templates test the radii; the girth tests their sum; the axes and tangent points test where the arcs meet; the flare and level test the cones between rim and bottom; and the capacity and leak tests prove that the vessel does its job. A trough that passes them all is the trough the numbers describe, and the numbers, from Chapter 2’s side radius to Chapter 9’s sheets, are what made it possible to check it at all.

Exercise: a tight side

Suppose a side piece were rolled too tight, to a radius of 1,400 mm. What would the side template, on its 1,000 mm chord, show? Would the trough’s girth reveal it?

Worked answer.

• A 1,400 mm radius rises 1,400 − √(1,400² − 500²) = 92.3 mm on a 1,000 mm chord, against the template’s 81.5 mm.
• The template would rock on the wall’s middle, with gaps of about 10.8 mm under its ends, well outside the 2 mm acceptance.
• The girth would probably not reveal it, because the side piece’s length along the rim is unchanged; pulled into the seams, the tight side bulges outward at its middle and the wall around the seams flattens to compensate, leaving the girth almost the same.

A tight side is as invisible to the tape as a flat one, and as obvious to the template. Both directions of error are why the templates are the first check on an oval, not the last.