A 50 mm Discrepancy on a 1:50 Layout and a Beam That Does Not Exist Yet
You are fabricating a rectangular duct run to a 1:50 layout. Along one wall the dimension chain reads 1200 + 1200 + 900 + 1200; the overall dimension on the same wall reads 4450. A general note states 'verify all dimensions on site before fabrication; do not scale from this drawing'. Your site measurements, taken between the finished faces of the wall at three heights, give 4480, 4486 and 4483 mm. The drawing also sets the duct centreline 300 mm off the face of a structural beam that has not yet been erected; the steel erector is due in three weeks and the erection tolerance on the steel drawings is ±25 mm. The wall is 100 mm blockwork laid to a string line and is not yet plastered. The duct is to be made in the workshop in 1.5 m sections and delivered flat-packed for assembly, with slip joints at every joint.
1. State which dimension governs, what you do about the three conflicting numbers, and what you must not do.
Read the worked answer
Hierarchy of information: site-verified dimensions override drawing dimensions; a specific dimension overrides a scaled measurement; and an overall dimension normally governs over a chain, because the chain accumulates setting-out and drafting error. Here the chain sums to 4500 mm against an overall of 4450 mm - a 50 mm internal contradiction that could only come from the drawing, since the site readings cluster around 4483 mm and are consistent with each other to 6 mm.
So the governing figure is the site measurement, taken conservatively. Use the smallest reading, 4480 mm, not the average. The wall is unplastered blockwork on a string line, so it will move slightly and plaster will reduce the clear space further; building to the mean or the largest reading risks a duct that will not fit once the plasterer has been in.
What you do:
- Record the three readings, their positions and the datum on a marked-up drawing.
- Raise a written RFI to the designer before cutting metal, stating the chain/overall contradiction, the site readings, and the dimension you intend to fabricate to.
- Have the answer issued as a controlled drawing revision, not a verbal instruction or a text message.
- Fabricate only the sections not affected by the disputed dimension until the revision lands.
What you must not do: scale the drawing (at 1:50, one millimetre on paper is 50 mm on site, and a 0.5 mm pencil line is 25 mm of duct), split the difference between 4450 and 4480, or average the site readings.
2. How would you handle the 300 mm offset to a beam that has not been erected, and what fabrication allowance would you build in?
Read the worked answer
You cannot fabricate to a face that does not exist. The 300 mm offset is dimensioned from a theoretical steel line, and the steel drawings themselves admit ±25 mm of erection tolerance - so the real beam face could be anywhere in a 50 mm band. Fabricating rigidly to the theoretical line has roughly a one-in-two chance of being wrong.
Practical approach:
- Obtain the steel drawings and confirm beam depth, camber and the erection tolerance, rather than working from the duct layout alone.
- Hold the sections that depend on the beam (the offset leg and anything downstream of it). Fabricate the upstream straight runs now, since they are unaffected.
- Build in adjustment at the interface: a purpose-made transition piece or adjustable sleeve with 50 mm of movement either way, or simply leave the last 1.5 m section to be measured and made after the steel is up. At 1:50 with a three-week lead time, this costs nothing except planning discipline.
- Agree with the designer whether the duct can be set off the structural grid line instead of the beam face, which removes the dependency entirely.
- Once the steel is erected, measure the actual face and issue a revised detail.
The judgement point is that slip joints are for assembly and thermal movement, not for absorbing 50 mm of set-out error - a slip joint forced that far will leak and pull apart under fan vibration. Adjustment must be built in deliberately, at a component designed for it.
3. What measurement technique and instrument would you use for the 4.4 m site check, and what errors would you be guarding against?
Read the worked answer
Instrument: a Class II steel tape, or better a laser distance meter cross-checked against the tape, measuring horizontally between the two finished faces at a consistent height.
Errors to control:
- Sag. A 5 m tape unsupported over 4.4 m drops in the middle and reads long - over a metre or so of unsupported span the error can be several millimetres. Support the tape, or measure in two hits from a temporary datum mark, or use the laser.
- Datum and reference face. 'Finished face' means plastered face. On unplastered blockwork you must add the plaster thickness, or you will fabricate oversize. Measure at the same height on both walls.
- Squareness and plumb. Two readings across a room can both be right and the room still out of square. Check the diagonal, and check the walls for plumb - a leaning wall reduces the usable width at duct level.
- Temperature. Negligible here: steel expands about 12 × 10⁻⁶ per °C, so 10 °C over 4.4 m is roughly 0.5 mm. Do not waste time on it.
- Tape-end and hook wear - check the tape against a known dimension, and always read from the same graduation.
Recording: log all three readings with position, height, date, datum and instrument, and the smallest value used for fabrication. That record is what protects you if the duct later does not fit.