Working with Errors

A bad works-man blames his tool - but the aim of this section is not to blame tool but to look for solutions. Errors are inevitable but can be minimised. Its the errors that sometime make work interesting and challenging. At this juncture, the project has not proceeded beyond documenting and measured drawing the built profile (plan). Hence this discussion is limited to what has been observed till this phase and includes :-

• Errors in establishing the Reference Plan
• Errors from Instruments used 
• Errors from Marking the Profile
• Errors from Measuring, Recording and Transferring dimensions 

Errors in establishing the Reference Plan

The positioning of reference plane is crucial for accurate location of a building. It needs to be exterior to the building, critically located so that determinant points along the section plane can be drafted from it. In this case, the closest possible permanent points from which a reference plane external to the building that could be generated was located 18.0 meters away. As a result, 4 points inside the building were chosen as the corners of the quadrilateral (reference plan) and measurements for interconnected spaces were taken with respect to the same. Over the length of 25.0 meters, it was observed that the cumulative error as a result has been 150 mm. 

After a discussion of the approach with Ar P. Latorre, it was realized that the process theoretically was correct, but in practice, the changes of errors, including human errors, area higher.  Also it was realised that drafting smaller areas from a larger reference plane negates errors than the other way round. The Diagram below shows the reference grid suggested by Ar P Latorre (Figure-1) and that executed at Site (Figure-2). 

Ar. Latorres recommendations were based on the possibility of external reference plane, which will be at this juncture a time consuming exercise being subjected to permission form the client. 

The consultants have now identified critical points (that include perfect corners, openings, significant niches or larger moulding) from which dimensions can be transferred to connected spaces. These points will be triangulated first and will serve as a reference grid for smaller interior details. The consultants have also in the meanwhile requested the clients to permit construction of bearing points within 3 meters from the built surface, which will serve as an external reference plane. 

On redrafting with the aforementioned approach, it has been found that the difference between the location of old and new points has been almost 40mm. 

Lesson learnt

• Incase an external reference grid is not feasible or cannot be availed - draft only  critical points to determine the overall profile of the section plane. The other details, like the niches, mouldings etc, in this case, should be drafted thereon. 
• The larger the grid the more correct it will be (perhaps).
• Avoid making moulding or other details as the critical points except when unavoidable. In this case, as there were higher number of chipped and or broken corners, one had to take edges of mouldings as reference corners. 

Errors from Instrument used

Knowing ones instrument is the first step to identifying possible errors and minimizing it. The consultants used the following combinations of instruments - first chalk to mark datum, nylon string to hold between 2 points on the plane and plumb bob, replaced by a spirit level to check horizontality of the marked datum line. Each set had its own limitation and implication on the errors in marking the datum line and thereon in drafting the profile of the section plane. 

Set one - A Plumb bob, String and Chalk - The plumb bob is probably one of the oldest tools at hand to check horizontal levels. Strings were held tort between two ends between which the level were to be checked with a plumb bob. Unfortunately, it was found that, due the wind, the plumb swung incessantly, leading to errors in reading. The use of this set of tools were abandoned in this case. 

However, if the distance between the points (where levels had to be checked) was lesser than 1 meter and the force of the wind was lower, perhaps the use of plumb to measure horizontality would have been effective. In case, it was in-effective. When verified using a spirit-level, considerable difference in level measuring up to 25 mm maximum was noted.

Set two - Spirit-Level, String and Chalk - A spirit level, at that point of time, was the most convenient replacement to the plumb-bob. Prior to procurement, the accuracy of the tool needs to be checked i.e., if the reference planes (top and side bases) are horizontal or not. In this case, while procurement, the tool was not verified for correctness and after commencement of work, the inclination of its reference planes were realized. 

Moreover, the spirit level gives correct reading within a certain range i.e. when the bubble remains within the 2 lines (Refer to Figure - Range within which reading of spirit level is considered correct). Since the reading remains affected by the judgement of the user, the  correctness of the horizontal level varies from person to person. 

Lesson Learnt

• It is a must to calibrate instrument and identity its errors before commencing work. Incase a large area should be covered, a smaller area must be measured as a sample to identify the extent and impact of errors from the instrument.                                                            As calibration is critical to accuracy, we plan to first locate the correction position of  the  bubble of spirit-level in use at this point to avert any subjectivity in reading. As a result of erroneous marking, there is a discrepancy of 80mm over an area of 25m long by 18m wide (Ground Floor).                                                                                                                                       

• Using a plumb bob is more difficult than it apparently seems, especially while using it for the first time. Hence, we feel that one must practice and verify the horizontality of the level acquired while marking with a plumb bob. Also, one needs to buy a plumb bob where the bob is heavy enough to negate wind action - a problem we faced. It is certain that plumb-bob will swing in windy areas. 
• A soft chalk is an inappropriate tool to mark as it leaves a thick mark as well as it fades.  It was initially chosen as the marks made would be temporary. However, the thickness of the chalk line as well as the mark it left on the building on erasing (incase the same had to be rectified) created confusion. Also, it was consciously decided that string or the measuring tape were to be held from the top portion of the marking (Refer to Figure - Point to Mark and Measure from). 

Errors from markings Profile

A - Cumulative Errors - This is possible the most critical of the errors and is cumulative in nature. It affects the horizontality and accuracy of the datum line and building profile. While marking the datum using a spirit level and the string, only 2 points at a time could be  marked. So the method used, was marking a second point from a previously marked point, eg - from 1, '2' is marked and from '2', '3' is marked and so on. Also, due to the nature of the built profile, rarely a series of point could be marked from a single point or the main the primary base point. As a result apart from the first set of points in a space, other dimensions were all transferred using a reference point. The series of marking increased the probability of a cumulative error. While visually surveying the profile,before measuring, errors in horizontality of the datum line were noticed. 

B- Nature of Built profile - The built profile in parts were inaccessible, broken or narrow, especially at the corners, making it difficult to even hold the end of the string. As a result,  it was difficult to mark the datum line along such corners or areas or points accurately, affecting the overall level.