# Snow Pack and Ice Loading Require Careful Study in Tower Designing.

Many articles have been written on the subject of snow loads, icing, cold temperatures and the like. It would take a good-sized textbook to completely cover the subject. In this article, i'd like to address a few of the areas based on my experiences, with the hope of enlightening other engineers and managers on some of the considerations of snow pack and icing design, as they apply to microwave towers and other structures.

Let's first look at what we should consider as snow pack. In many regions of the north, and even in mountainous regions of the south, there is a considerable amount of snowfall. In many places this can amount to a mid-winter depth of 15 to 20 feet. This snow pack can last for several months, and poses a definite challenge to the structural engineer.

Depending on elevation, latitude, location, prevailing winds and other conditions, this snow can either by dry and relatively light in weight (about 30 pounds per cubic foot) or it can be very wet and heavy (about 50 pounds per cubic foot). With water wet and heavy (about 50 pounds per cubic foot). With water having a density of 62.5 pounds per cubic foot, it is easy to see the amount of weight actually being considered. These ice weights are consistent with the Electronic Industries Association standard RS-222-C, "Structural Standards for Steel Antenna Tower and Supporting Structures," which considers solid ice to weight 56 pounds per cubic foot and "rime" ice to weight 30 pounds per cubic foot.

When estimating the weight of the snow at a particular site, don't be fooled by the outward appearance. As the seasons change, temperatures go up and down, and storms come and go. The snow is deposited in layers, with each layer melting on the top and compressing the layers below. This consolidation of snow layers adds greatly to the pressures below it, simply by increasing the weight. In addition, since water is heavier than snow, it will tend to seek the lower snow-level layers, adding to the density of the lower layers. This is where the term "snow pack" comes from; the snow is actually packing down.

Any structure that is to be installed in this type of environment must be designed to resist these loads. It is necessary to determine a method for converting the weight of snow in pounds per cubic foot into a uniformly applied load in pounds per lineal foot, PLF. One method is to estimate the snow depth above the member being designed and multiply this by the width of the member. As an example, consider a four-inch-wide member with eight feet of snow above it. Estimate the snow weight at 50 pounds per cubic foot.

snow volume = (4/12)(8')1') = 2.67 cubic feet per foot of member

snow weight = (2.67)(50) = 133 PLF

Therefore, use 133 PLF for uniform design load. Since the contact area is 0.33 square feet, this load converts to 399 PSF. This shows the potential magnitude of the loads to consider.

In many cases a structure may be located on a relatively flat area of a hillside, often on an area cleared and leveled especially for the structure. This can create problems for a structure by placing it in a location subject to snow sliding and drifting.

On most hillsides any large accumulation of snow will tend to slide downhill as the winter progresses and as additional snow is deposited on the top. If the hillside is heavily covered with trees, it can help to hold back the snow. On very steep hillsides, this snow sliding and packing against anything in its way can add considerable amounts of weight to any structure. The effect of this can be seen on most any mountaintop (such as local ski areas) in the way trees are bent and pushed downhill.

If the structure is designed of steel we now have all three ingredients necessary to design the member for combined axial compression and bending, per the requirements of the AISC (American Institute of Steel Construction) manual of steel construction, Section 1.6.1.

It can be shown by the principles of statics that the horizontal load to apply is approximately equal to the weight of the snow on a member times the sine of the angle of the slope of the hill, from horizontal. Therefore, for a hill slope of 30 degrees the sine of 30 degrees is 0.5, and the horizontal load to apply should be 0.5 times the snow weight in pounds per lineaf foot (PLF) or pounds per square foot (PSF), in the case of using applied pressure.

It has been observed many times that a site that appears to be a natural tower or passive repeater site can turn out to be a very bad site in terms of snow loading. Engineers have known for a long time that snow will creep with time. As large amounts of snow creeps, it usually takes with it small trees and most anything else in its path. Now think back to why that site you were considering looks so perfect. It may be that the severe snow load and creep has literally cleaned everything off the hill.

If a structure is built on this site it certainly will not stop the snow from creeping down the hill; it will just provide a place for the snow to build up. Several years ago I worked on a passive steel channel providing full coverage over the waveguide runs. This type of ice shield can also be used to cover the waveguide between the antenna and the tower.

Additional considerations should be given to protecting the electrical power lines that enter a communication building if they are in the proximity of the tower. This is one reason that many sites have underground service entrances.

Ice falling from a tall tower can also hit the top of the communication building, severe engough to cause structural damage or at least cause the roof to leak. Most portable buildings are designed for extreme roof loading to protect the structure. They may also be provided with a shield covering the entire roof. These are usually built with bar grating and fastened directly to the building walls to transfer any loads directly to the building foundations and not allow any to be transmitted through the roof.

The amount of damage that can be caused by falling ice should not be underestimated.

As an aid to designers, most states publish what is referred to as a "snow-load analysis." This is usually published by the structural engineers' association of each state. These books are handy for planning the ammount of snow load to expect at a particular area. The information is usually broken down by county and city and can be present in a tabular form giving minimum ground elevation, actual elevations of major towns, expected ground snow load and recommended roof snow load in pounds per square foot (PSF).

Other good information that can be found in these books is a brief description of the physical properties of snow and information on potential drifting problems. These books are mainly directed toward roof designs, but they provide a good source of information on the potential snow load that a particular town may experience, to aid the engineer in prepaing for possible snow-pack conditions.

Earlier, I discussed a few of the potential problem areas that can cause considerable ice and snow loads on microwave towers and passive repeaters. Now, I will show a number of ways that the engineer can guard against structural damage and insure that the tower will have a long service life even in the most severe winter areas.

Since snow and ice like to build up around every member, it is a much better policy to design a structure with with a few very large members instead of a lot of small ones. It is not good engineering to try to consider just the weight of steel in a structure to determine if it is cost-effective design to resist snow-pack loads.

If a structure has a large number of members, the collecting snow and ice will eventually "grow" large enough to completely fill in the open spaces between members. At this point the structure will then block the blowing and drifting snow and build up a tremendous amount of pressure, with only relatively small members to resist the loads.

If a structure has just a few large members it is not as likely that all of the open spaces between members will be albe to fill in solidly with ice, thus allowing some of the snow and ice to continue to blow through the openings. In this way, the total load on the structure is a lot less, and the resisting members are relatively large.

In the design of structures to resist snow-pack loading it is worthwhile to consider members that have the same section modulus (S) in both horizontal and vertical directions, such as pipe or square tubing. This assures the same bending strength in both directions. Some structural shapes have additional problems to consider, such as "local web buckling" of pairs of double angles. With double angles the buckling of outside flanges (outstanding legs) is the controlling factor and is they way most bending failures begin. Watch Inside Buildup

Structural pipe is an excellent section to resist combined axial and bending stresses due to its constant "S" in all directions, its reduced wind loading (due to the round shape) and its excellent strength to weight ratio. Its biggest draw-back, one that requires careful engineering, is the potential of snow and ice buildup on the inside. If allowed to build up and freeze, this ice can expand and cause the pipe to crack. This can be avoided by designing members that are closed on the top, with weep holes on the bottom to allow any moisture that collects on the inside to drain properly, or by designing completely open members to allow full drainage.

If a member is subjected to only axial loads, the orientation of the piece is not critical to its load-carrying capacity. In the case of members designed to resist combined axial and bending stresses, it is very important to orient the piece properly.

Since the bending stresses are directly related to the section modulus, the member should be oriented with the greater section modulus in the direction of highest expected loads. As an example, in wood design if a single 4 by 10 is located with the 10-inch nominal dimension in modulus is 49.9 in \$(n3\$), and the horizontal section modulus is 18.9 in \$(n\$). This shows that the member will support nearly three times as much load in the vertical direction as in the horizontal direction.

An additional problem comes up if this member is used on a hillside where a horizontal load is to be applied. On a hill slope of 30 degrees, the applied horizontal load is half of the applied vertical load, but the contact area (for a 4 by 10) is 9.25 inches divided by 3.5 inches, or 2.6 times the vertical contact area. This gives an applied load of 30 percent more in the horizontal direction, due to the larger surface area, but a considerably weaker section in this direction.

This same analogy can be used for steel members, where it is also necessary to locate the direction of hightest section modulus in the direction of highest applied loads.

There are many excellent texts on the subject of composite beam designs, so I will not attempt to duplicate this effort, only to point out the need to consider composite beams.

In the case of combined stresses it is often necessary to design a composite beam to put the maximum section modulus in the direction of maximum applied load. If designing for snow pack it becomes a trial-and-error process to come up with an adequate and economical design. As additional members are added to the base member to increase the effective section, you also add surface area that can increase the applied loads, thus requiring another check of the bending stresses.

It is important to consider the consquences of designing a particular built-up member. Consider the case of a horizontal member composed of a pair of double angles, as shown in Figure 5.

A check of the applied loads in the horizontal direction shows a need to increase the section modulus, "Sy." This is easy to do by adding a horizontal plate to the top flange as shown in Figure 6.

This may seem to have solved the problem except for two important considerations:

Adding a plate requires a considerable number of bolts to make the section effective. AISC section 1.18.2.3 requires that for ASTM-A36 material and three-eighths-inch-thick plate, the hole spacing should not exceed eight inches. This would require 28 bolts in a single 10-foot-long member. This would take an excessive amount of fabrication and assembly time, in addition to the cost of the hardware.

The plate adds very little to the capacity of the member in the vertical direction while actually increasing the member weight and the corresponding bending stresses.

A more efficient and cost-effective solution would be to add an additional pair of angles to the top flange, as shown in Figure 7.

This increases the section modulus in both vertical and horizontal directions, and requires far fewer connecting bolts than a plate. Since adding section modulus to the vertical direction, it may be possible to select a smaller-size angle than previously thought.

in designing a structure to resist severe ice and snow loads it is not always economical to merely increase the member sizes until attaining adequate strength. At some point it may be a good idea to consider high-strength steel.

Most steel used in communications structures today is ASTM A36 grade, having a yield strength of 36,000 PSI. This is the grade of steel used in angle iron, plates and other structural shapes. Most pipe is ASTM A53, Grade B, with a yield strength of 35,000 PSI. Since both of these grades have approximately the same yield strength, the American Institute of Steel Construction (AISC) allows engineers to use the same allowables. Choosing Grade of Steel

If it is decided to use a higher-strength steel, what grade should be used? In the front of the AISC Manual of Steel Construction is a table listing the major grades of steel and the corresponding shapes that most steel mills produce. It is absolutely necessary to check with local steel warehouses and mills for availability before considering a certain grade in your design.

One of the most readily available grades of high-strength steel is ASTM A441, Grade 50, with a yield strength of 50,000 PSI. This is available in most common structural shapes and plates. Pipe is also available in high-strength steel, but availability of any one grade is limited. Pipe can, at times, be found in ASTM Grade A333, Grade 9, with a yield strength of 46,000 PSI, and tubing can be ASTM A500, Grade B, yield strength also 46,000 PSI. Another consideration for structural shapes is ASTM A572, Grade 50, having a yield strength of 50,000 PSI.

The benefit of high-strength steel is in its higher bending strength allowable, since nearly all failures due to snow and ice loading are of a bending nature. Since the allowable bending strength is in direct proportion to the steel yield strength, the higher the yield strength, the higher the bending allowable. AISC gives the allowable bending strength for structural shapes of 0.66F \$(y\$) (0.66 times the yield strength), therefore, 50,000 PSI steel is 39 percent stronger in bending than 36,000-PSI steel.

The additional cost for high-strength steel is more than offset by additional strength and in keeping the structure weight down. Since most shipping is by the pound, the added cost of raw steel can be offset by reduced shipping charges. As an added thought, the installation crews will surely thank you for cutting down on the weight of the pieces that they may have to drag up some lonely mountain peak.

As stated previously, it is much more efficient to use a small number of large pieces instead of a large number of small pieces. In trying to accomplish this, some of the pieces can get to be too long to be efficent to resist bending loads. This is where the high-strength steel comes into play. In the basic formula for bending moment for a beam with pinned ends (as in the case of a single bolted connection), the bending moment increases as the square of the length: M = wl 2/8 where M = bending moment in pound-inches, w = ice + member weight uniform loading in pounds per inch of member length, l = member length in inches. Hence, the longer the member, the more critical the bending strength, and the more appropriate it may be to select high-strength steel.

In the design of structures to resist axial and bending stresses, often the most economical section can be double angles. Double angles consist of two pieces of angle iron, equal or unequal leg, bolted together to act as a single unit (see Figure 5). The bolts used to bolt the angles together are referred to as "stitch bolts" and include a spacer of the proper thickness to match the connecting end-plate thickness.

For double angle members designed for axial loading, it is necessary to space the stitch bolts so that the l/r (the member length divided by the member radius of gyration) of the single angles member is less than the l/r of the double-angle member (see Figure 8). The l/r of the single member is checked in the Z-Z direction, its weakest direction, and is compared to the l/r of the double member of the Y-Y or X-X direction. See the AISC Steel Construction Manual for additional information.

For relatively short members, it is possible to get by with a single stitch bolt in the center, halfway between each end-connecting bolt. For members designed to resist snow pack or radial ice (bending stress), this is the worst possible location for a stitch bolt. Placing a stitch bolt in the center puts it at the location of maximum bending moment, which results in the highest tension in the extreme bottom fibers (see Figure 9).

Adding a stitch bolt requires making a hole in each angle, thus removing a large amount of cross-sectional area. Being able to physically place the bolt in the angle and therefore doesn't leave very much edge distance (see Figure 9). All of the material in this edge distance will be in tension, due to the bending loads, plus any axial tension loads. In many cases this edge-distance material will not be sufficient and the member will fail in tension at the stitch bolt hole.

To check this it is necessary to calculate the resultant "section modulus" with the bolt hole section removed. Then the bending stresses at the bolt hole can be checked to see if the section is adequate. A Good Policy

For the cost of an additional stitch bolt it is a good policy to always put at least two stitch bolts in every double angle member subjected to bending stresses. These should be located at approximately one-third the member length.

The main points are to keep stitch bolts away from areas of maximum moment (the mid-span of a simply supported beam), to put in at least two stitch bolts, to keep the stitch bolts (and consequently the bolt hole) as small as possible and to maintain maximum edge distance on the tension side.

Along the same thinking is the problem associated with connecting one piece of angle to a pair of double angles at the center of the double angle pairs, as shown in Figure 10.

This is a typical situation of how secondary members are connected to main members. The problem created is the same as locating a stitch bolt in the center.

If designing for snow-pack loads, a better connection would be to use a plate as shown in Figure 11. This removes the potential of having the main double angle pair fail in tension at the center bolt location.

A consideration for controlling the bending moment is the type of end connections used on each member. As shown before, the bending moment is the controlling load for snow pack, and for a pinned (single bolt) end connection the bending moment is M = wl 2/8. For a beam with the ends fixed, as in a welded connection or multiple bolted connection, the bending moment is M = wl 2/12. From this comparison it is easy to see the importance of providing a rigid end connection, as it can reduce the bending moment by one-third.

When designing members to resist snow/ice load, it is indeed foresight to require all main load-carrying members to be at least double-bolted (two bolts at each end). This greatly reduces the effective length and therefore the bending moment of each member, and hence increases the load-carrying capacity. This comparison (see Figure 12) assumes, of course, that the connecting end supports will remain relatively rigid.

Many structures designed for use in arctic environments should be painted black to take full advantage of the sun's rays, thus causing a quicker melting of the snow and ice.

Even if only a portion of the structure that is black is exposed to the sunlight, this part will heat up, and the member will conduct this heat to other parts of the structure. The biggest advantages are the speeding up of the snow and ice melting in the spring to help relieve the structure of its snow-pack loads, and to cut down on the amount of ice buildup on a member by melting al little of it very time the sun comes out.

Due to the conduction of heat through a member, it will tend to melt the snow directly in contact with it and will leave a "cylinder" shaped opening in the snow, thus releasing some of the pressure from the member. This does not always release all of the pressure, and in the case of a horizontal member it will melt the snow in all directions, but on the top surface the snow will just continue to pack down into the melted area. This can leave a structure with a large amount of snow melted from under the members, with the members still carrying the weight of all the snow above them, as shown in Figure 13.

This "cylinder melting" effect shows up on all structures but is much more pronunced on those structures that are painted black.

Repairing a structure damaged from high snow and ice loading is an extremely expensive and frustrating project.

If a structure is in a position to receive high snow loads, more than likely it is also at a site with very difficult access, ofthen where a helicopter is the only method of transportation. It's easy to imagine the cost of rehabilitating a damaged structure, or in "beefing-up" an existing structure, could approach or exceed the cost of the original structure. Expensive, yes, but always less than replacing a tower that was overloaded to the point of complete collapse.

If an existing tower is considered for a major strengthening project, it should be looked at very carefully by a competent tower engineer. It's easy to say that a tower should be strengthened, but it is virtually impossible on some towers. Making an existing tower stronger is a snowballing process (pun intended). Adding more members or bigger members adds even more wind, weight and snow load to the tower. This added load must be carried through the entire structure and every member checked.

It is feasible to change out horizontal and diagonal members with larger ones but very difficult, if not impossible, to change out a tower leg. About the only thing that can be done to increase the compressive strength of a tower leg is to add additional secondary members to shorten the effective length of the leg, as shown in Figure 14.

It takes very little imagination to visualize what can happen to a tower when a member is removed to be replaced with a larger one. The structure becomes very unstable. Tower engineers have heard more stories than they would like to admit of tower failures due to inexperienced workers removing the wrong member at the wrong time. This type of construction should be attempted by only very experienced tower crews, under under direct supervision of the tower engineer.

If a structure has been damaged due to heavy snow loads, it is necessary to try to anticipate the actual loads that may have caused the damage, so that the proper size pieces can be provided. Field-Check Lengths

When replacing members of an existing structure it is a good idea to field-check all member lengths. A structure that has been up for a few years will take on a certain set due to load reversals from the same direction year after year, and the tower may actually have fabrication errors that were built into the tower. When a tower is first built, all connections are left loose during construction to make it easier to install the various pieces and make up for any differences in member lengths. If a tower is already up it is very difficult to get enough slack in the members to allow the wrong length member to be wedged into proper position. Since the tower is rigid it's difficult to make up a difference of a half-inch in a member length that won't fit between a fixed set of end holes.

It should be clear by now that microwave communications towers need to be survive extreme winter weather. Microwave towers by their very nature must often be used in large mountain ranges to enable microwave signals to be transmitted across these mountains. The possibilities of heavy snow pack are very high and should be considered at the very beginning of the project. Likewise, some parts of the country do not have heavy snow loads but have extreme icing conditions.