A problem has come to light that may cause wind loads on lattice towers with large ancillaries to be calculated incorrectly when Mstower is configured for wind loading to BS 8100 or AS 3995. The problem could affect any job with a large ancillary whose reference point is inside the tower faces, e.g. platforms, cabins, radomes, etc., largely inside the tower. Ancillaries above the tower are not affected.Mstower V6.20.02.04 (referred to below as build #16), is the first version containing a fix for the problem. It may be downloaded from the Bentley Software Fulfillment Center website.
The way Mstower calculates wind resistance of lattice towers with ancillaries is described in the manual:
MStower uses the general method of BS 8100 for computing the wind resistance of towers. This method allows for towers with faces that are asymmetrical, either structurally or due to their complement of ancillaries. It also allows the resistance to be computed for any wind incidence angle. When using the general method, the resistance of the single frame comprised in each face is computed, along with shielding factors and Kth. The resistance of the complete tower is built up from these values. Methods of computing drag coefficients of panels made of flat and circular sections (both sub-critical and super-critical) are also given. BS 8100 also uses a simpler method for symmetrical towers, whereby the resistance for the complete tower can be determined from drag factors for the overall tower.
If a panel contains ancillaries, the projected area of the ancillary is used when computing panel solidity ratios and single panel drag coefficients. The wind forces on the ancillary are then computed using the drag coefficients from the ancillary library and a statically equivalent set of node loads is applied to the nodes to which the ancillary is attached.
When AS 3995 is specified MStower uses the general method as described above but with single frame drag coefficients that give overall drag coefficients equal to those in Table 188.8.131.52 of AS 3995. This allows the program to maintain the ability to deal with towers that are asymmetrical or composed of mixed section shapes. It also allows wind forces to be computed for angles of incidence other than face and corner. For a tower carrying large dishes, the critical wind may occur at some other angle, which may vary from member to member.
Mstower allocates large ancillaries to one or more faces depending on the coordinates of the reference point. If the reference point is located more than 95% of the distance (measured horizontally at the ancillary level) from the tower axis to any face the large ancillary is allocated to that face.
When configured for wind loading to BS 8100 or AS 3995, Mstower has always used the method of BS 8100 Part 1 Sec. 4.4 “General method for towers containing ancillaries or unsymmetrical towers” to compute the combined resistance of tower and ancillaries. This method involves adding ancillary resistance to the resistance of the equivalent bare tower. Panel solidity ratios are computed using the projected areas of the structural components and large ancillaries assigned to the face(s), and these are used to compute the resistance of the equivalent bare tower.
Linear and face ancillaries are handled by Mstower in much the same way as large ancillaries. Linear ancillaries allocated to a face will affect the panel solidity ratio while those not allocated to a face will not. Face ancillaries are always associated with a face so the corresponding panel solidity ratios will be affected.
There is a bug in all versions of Mstower up to V6.20.01.15 (build #15). The effect of the bug is to underestimate the resistance of lattice towers with large internal ancillaries. This occurs because the wind resistance of large ancillary items used to calculate equivalent bare tower resistance contains a contribution from large internal ancillaries that may not be warranted. This bug has been fixed in version V6.20.02.04 (build #16). This version shows the notation “not on face” for any large ancillary inside the 95% threshold.
Large internal ancillaries include platforms, cabins, radomes, or similar that are input to Mstower as large ancillaries. These are not allocated to any face when the reference point is inside the 95% threshold line. If you want to ensure that a large internal ancillary is to be treated as an ordinary large ancillary it may be necessary to replace it with multiple large ancillaries, one for each side. With a platform extending outside the tower faces, for example, as a single ancillary its reference point would be inside the 95% threshold and its resistance would not affect the calculation of the equivalent bare tower resistance. As three or four separate large ancillaries each of the sides would have a reference point outside the 95% threshold and would affect the calculation of the equivalent bare tower resistance.
To model large internal ancillaries that do not affect the panel solidity ratio it may be reasonable to use a SHADE factor less than 1. This is never done for a large ancillary on the tower face but it should be acceptable for a large internal ancillary that is shielded to some extent by the tower itself.
There is a keyword SEP that may be added to any large ancillary line in the .TWR file, which forces the large ancillary resistance to be computed separately. When the SEP keyword is present for a large ancillary its resistance will not contribute to the resistance of large ancillaries used to calculate the equivalent bare tower resistance. For a tower without linear or face ancillaries, using the SEP keyword for all large ancillaries will produce the result obtained using BS 8100 Part 1 Sec. 4.3 “Method for symmetrical towers with limited ancillaries”.
Install release 6.20.02.04 or later.