Sunday, January 15, 2017

Hope and Wind Are 4-letter Words

Last week we had a wind storm that provided the first test of the survivability of the new tower and antennas. While everything did survive there was an effect that calls attention to the danger of assumptions and the best laid plans. Worse is relying on hope alone, which is all too common in the amateur ranks.

I didn't do this yet I did fall victim to my assumptions and a "sensible" risk assessment. As with hope, wind is also a 4-letter word. (For those outside the North American English language culture, many of the common obscene/swear words are 4-letters long.) There are no pictures of the antennas in this article. It would not be informative since a picture would only show the tower and antennas standing there as usual. A picture would not show what happened.

The wind

This was not a severe weather event. For many it was welcome since it brought a strong southerly air flow that quickly raised the temperature to a near record high. The winds were sustained at and above 50 to 60 kph for almost 12 hours with gusts up to and possibly exceeding 90 kph. A better estimate isn't possible since I am not near a weather station and the darkness did not allow for direct observation. Instead I slept, reasonably secure in the confidence of my planning.

The wind was within my survivability threshold even with the relatively large wind load I have temporarily placed atop the Trylon. Since the wind area is at its lowest with the yagis broadside to the wind I turned the yagis northwest at the start of the storm. The wind was from the southwest. In the morning I discovered that they were pointing northeast, with the booms aligned with the wind. That is, the mast slipped within the Tailtwister mast clamp until the maximum wind area was exposed to the storm wind.

Precautions taken

When I installed the mast and rotator I deliberately avoided pinning the mast to the rotator. I didn't even excessively torque the u-bolts of the clamp. In consideration of the antenna wind area my preference is that if the yagis want to move I'll let them. I'd rather that than have the peak torque transferred to the tower. Climbing the tower and turning the mast is preferable to risking the integrity of the tower.

Some hams don't like the idea of climbing, or depending on others to do it for them, and prefer to pin the mast. A safer approach is to wind balance the yagis, whether home brew or commercial product. There are software tools available to do this, or one can follow the construction details of yagis that have been wind balanced, such as those in the ARRL Antenna Book. My yagis are moderately well balanced, however I have not confirmed that.

Calculation of antenna wind area

Calculating the wind area of elements can be time consuming and prone to error. There are many factors: taper schedule, boom-to-element clamps, tubing clamps, coils, traps, capacity hats, linear loading, matching networks, etc. Most just go by what the manufacturer claims in the product specification. This can be misleading since some report the highest wind area (dependent on antenna orientation to the wind), lowest wind area or something in between.

Booms are more easily calculated. There's just the aluminum tubing, clamps and a small protrusions such as beta matches, baluns and trusses. However it is more typical that the elements have a greater area than the boom so we need to know that area. It is also helpful to know that the maximum wind area is when either the boom or the elements point directly into the wind, not at some point in between. I refer you to the excellent work of K5IU and the cross flow principle in Spring 1993 Communications Quarterly (I can't locate a link right now). Yagi design tools such as YagiStress now include this work.

I finally decided to buckle down and do the calculation with a calculator in one hand and the XM240 and Explorer 14 manuals in the other. I'll spare you the messy details. Even with all the calculation there will be some error in the true wind force since antenna elements with many diameter transitions are imperfectly modelled by the long cylinder wind load coefficient.
  • XM240 boom: 5.5 ft², including estimates for the balun, boom-to-mast clamp and boom truss.
  • XM240 elements: 3.7 ft², including an allowance for element-to-boom clamps, tubing clamps and the capacity hats.
  • Explorer 14 boom: 3.5 ft², including estimates for the boom-to-mast clamp, beta match and coaxial choke.
  • Explorer 14 elements: 10.2 ft², including an allowance for element-to-boom clamps and tubing clamps.
The wind area of the tri-bander elements is higher than I expected and is more than the 7.5 ft² I found in my copy of the Hy-gain manual. They appear to be using the average of the boom and elements. That is misleading and potentially dangerous. Cushcraft does better, specifying 5.5 ft² which agrees with the maximum value that I calculated.

Many hams who stack yagis are aware that tri-bander elements have a large wind area and choose to mount those close to the tower top plate and an XM240 or similar antenna above it. I know it as well but chose to do the opposite for expediency considering the weather difficulties of winter antenna work. Getting the heavier antenna higher and the even higher boom truss is not easy in our winter weather.

I performed my survivability calculations accordingly, but with the published wind load value. That was a mistake even though I know to turn the booms broadside to the wind. Recalculating for the yagis pointed into the wind the mast survivability dips to below 100 kph. That's quite poor, although I was conservative in choosing the steel strength estimate for the mast. In fact the range of wind survivability is 30+ kph, depending on the wind direction.

Shadowing

As wind speed increases there is increased shadowing for close spaced element. This is due to the time it takes for the airflow to recombine after being diverted around an antenna element. Laminar flow is not immediately reestablished, instead resulting in turbulence.

The only likely location of shadowing in my installation involved the so-called para-sleeve (coupled resonator) surrounding the Explorer 14 driven element. If there is shadowing the wind area reduction in high winds is probably no more than 1 ft². That isn't helpful.

Conclusion

The tower survivability remains sufficient even with the revised wind area calculation. The mast however is more of a concern than expected. All I can do for the time being is to lower the yagis a small amount and gain 10 kph of wind survivability. The XM240 can go down 6" so the Explorer 14 could go down a foot or more and only risk added 15 meter pattern distortion. With 15 meters so poor at this point in the solar cycle that may be the right thing to do. It's only until spring.

I could also increase the mast clamping force to reduce the chance of the mast turning in a high wind since weather vaning seems to be a problem. Or the clamp could have an added insert that bites into the metal rather than solely rely on friction. But pinning the mast is a step too far in my opinion. A few days after the storm I reoriented the yagis, put a little more torque on the rotator mast clamp bolts and tightened some other fasteners I had suspicions about.

While I am not terribly concerned about the mast it would be inconvenient if it were to bend. Hopefully statistics and bolt torque are enough protection. Otherwise I'll need to take action or accept a higher risk until spring.

No comments:

Post a Comment

All comments are moderated, and should appear within one day of submission.