Galicia PAU Technical Drawing Applied to Plastic Arts and Design II common mistakes
Galicia PAU Technical Drawing Applied to Plastic Arts and Design II · Understand the traps and how to avoid them.
Drawing the tangent line or arc between two circles by eye and marking the tangency points approximately, without constructing the perpendicular from the centre to the line of centres or using the radical-axis construction.
Candidates remember the picture of a common tangent but not the locus-based method; under time pressure they 'close' the figure by hand and the tangency point drifts off the true point.
What to do instead
Before inking, construct the auxiliary circle of radius R1−R2 (external tangent) or R1+R2 (internal tangent), draw the tangent from the second centre to it, then drop the perpendicular to recover the exact tangency points; verify by checking that the radius at each tangency point is perpendicular to the tangent line.
Scaling a drawing by measuring a distance with the ruler and multiplying it, instead of using the graphic scale, and writing the resulting dimension rounded to a convenient number.
The candidate treats the scale as a mental arithmetic step and forgets that the drawing is a reduced model; the printed graphic scale is never used and accumulated rounding is invisible.
What to do instead
Always read magnitudes off the graphic scale of the opposite scale (or the counter-scale) with dividers, and write the true dimension as the measured value times the scale denominator; if a value must be rounded, round to the precision the scale can actually resolve and state the unit.
Dimensioning with chains of dimensions from feature to feature so that the total length is also given, producing a closed chain that over-constrains the part.
Candidates dimension each feature as they draw it, left to right, and add the overall length 'for clarity', which is the classic redundancy error.
What to do instead
Choose a datum (an edge, a centre line, an axis) and dimension every feature from that datum in parallel; leave one dimension of the chain open, and give the overall dimension only when it is the functional one, never both.
In dihedral projection, drawing the auxiliary view of a slanted face by projecting perpendicular to the folding line instead of perpendicular to the true-shape plane, so the auxiliary view is not a true shape.
The candidate applies the rule 'project perpendicular to the fold' mechanically without identifying which plane the new view must be parallel to; the direction of projection and the direction of sight are confused.
What to do instead
State the target first: to see a slanted face in true shape, the new projection plane must be parallel to that face, so the direction of sight is perpendicular to it; set the new folding line parallel to the edge view of the face, then project all points perpendicular to that new fold.
Mixing first-angle and third-angle symbol conventions within one sheet, placing the plan above or below the elevation inconsistently and omitting the projection symbol.
Software templates and textbook figures come from different traditions; the candidate copies whichever layout is at hand and never checks the symbol that defines the method.
What to do instead
Decide the method at the start of the sheet, draw the corresponding truncated-cone symbol (first angle: cone with the small end toward the viewer, plan below the elevation; third angle: plan above), and keep every view of every object on the sheet in that same arrangement.
Building an isometric or cavalier axonometric by measuring true lengths along all three axes, including the receding axis, and then wondering why the object looks stretched.
The candidate confuses the axonometric scale with the natural scale and does not know the reduction coefficients; the receding axis is treated like the other two.
What to do instead
Use the standard coefficients: in isometric all three axes are reduced by 0.816 and drawn at the same scale (the isometric scale), or work on the isometric grid where equal divisions on the three axes are accepted; in cavalier the receding axis is halved, in cabinet it is also halved and drawn at 45°.
In a one- or two-point perspective, placing the station point, horizon and measuring points arbitrarily, so verticals tilt, the picture looks distorted, and the vanishing points fall off the sheet.
Candidates draw the perspective from imagination and do not link the station point distance to the cone of vision; the 60° cone and the 90° angle at the station point are not used.
What to do instead
Fix the horizon and the picture plane first, place the station point so the object falls inside a 60° cone of vision, locate the vanishing points from the angle the object's edges make with the picture plane, and use measuring points (or the diagonal method) for depths rather than guessing.
Choosing type size, line spacing and alignment by eye on screen at 100 % zoom, so the typographic hierarchy collapses when the piece is printed or viewed at real size.
The candidate designs on a zoomed canvas and never checks the actual reproduction size; hierarchy is judged by how 'big' the letters look on the monitor rather than by the reading distance and the format.
What to do instead
Set the format and reading distance first, define an explicit type scale (for example a modular ratio between body, subhead and headline), set leading as a multiple of the type size, and proof at 100 % and on paper before approving the hierarchy.
Specifying colour only as an RGB or hex value, or judging a palette on an uncalibrated monitor, so the printed result and the accessible contrast differ from what was approved.
The candidate works entirely in a screen workflow and does not distinguish additive from subtractive mixing, gamut limits or the need for a spot/CMYK specification.
What to do instead
Carry both an RGB value for screen and a CMYK or spot specification for print, check the spot colour against a physical guide, verify text/background contrast against the WCAG ratio, and never approve a palette without a calibrated screen or a printed proof.
Submitting the CAD file with layers named 'Layer1', unmerged duplicate lines, dimensions exploded into lines and no plotted scale, so the drawing cannot be read or edited by another person.
The candidate uses CAD as an electronic pencil and never learns the organisational layer of the tool; the deliverable is judged only by how it looks on screen.
What to do instead
Adopt a layer convention before drawing (name, colour, lineweight per element type), keep dimensions associative, purge and audit the file, and plot from a named layout at a stated scale with a title block; check the PDF at 1:1 against the graphic scale.