Ever wondered how to choose the tubing to use out the top of your tower?

here is some information that might help you decide

3rd Feb-2009, Guy vk2ku asks:

This is a question for the engineers among us!

Until now I have used 48.4mm diameter aluminium scaffold pole, 4.47mm wall thickness, for my antenna mast on a Nally tilt-over tower. Aluminium temper of this material is 6061 T6.
With a 4x12 array on top of this, I have experienced no problems, even when tilting the tower to horizontal, though I do support the mast once it is tilted horizontal.

I am now contemplating a new, bigger, and heavier array on a similar (but single-section) tilt-over Nally tower. Both the thrust bearing on top of the tower and also the rotator itself (2.4m below the top) will accommodate a mast of diameter 64mm (2.5 inches) for greater strength.

Q1. What material should I choose for the mast? Capral list a 63.5mm diameter extruded aluminium tube with 6.35mm wall, but the temper is 6106 T6. Might be ok?
Weight of 5m of this tubing would be about 15kg. Alternatively I could use some kind of steel alloy - what would be best, and what wall thickness should I go for?

Q2. Whatever I choose for the mast, I would prefer to carry the vertical load on the thrust bearing rather than on the rotator below.
That means some kind of sleeve on the mast which will sit on the thrust bearing.
How should I attach the sleeve so as not to weaken the mast at this critical point?
Bolting it through with 2 cross-bolts will surely weaken it.

Advice from experienced engineers would be appreciated.



3rd Feb 2009, Peter vk3qi replies:

Hi Guy,

Your answer needs several parts.

PART 1: The yield strengths are as follows:

Material Alloy MPA PSI
Al 6061-T6 241 35,000
Al 6061-T5 (over 25 mm thick) 110 16,000
Al 6351-T5 (known as scaffold tune) 250 36,000
Al 6351-T6 285 41,300

As a general rule, the extruded tube 48.4 or 63.50 that you talk about is 6351-T5, so you work on a figure of 250MPa or 37,000 psi

The yield strengths for steel tubing GENERALLY available in Australia are:
Material Alloy MPA PSI
FE C350 (used to be called schedule 80) 350 50,000
FE C250 (used to be called schedule 40) 250 36,000

You can sometimes get Chrome Moly which is 530 MPa = 75,000 psi

So ALL OTHER THINGS BEING EQUAL, using Fe 250 or Al scaffold tube gives virtually the same yield strength.

By "all other things being equal" means the internal and external radii of the tube are the same and the length of the pole load is the same

PART 2: To work out the actual bending moment that a particular tube can withstand, you need to know three things

(1) Outside radius
(2) Inside radius
(3) Yield strength

The actual formula is:     [R(outer)^4 - R(inner)^4] times yield strength

The calculations can become tedious, so it is often easier to work out by comparing with a known situation such as your existing scaffold tube:

48.4mm diameter aluminium scaffold pole (Al 6063-T5) with 4.47mm wall thickness

    [(24.2^4 - 19.73^4)] x 250 = 191441 x 250 = 47,860,250    (Al 6063-T5)

Compared with, Al 63.5mm OD 6.0mm wall thicknes 6063-T5

    [(31.75^4 - 25.75^4)] x  250 = 576538 x 250 = 144,134,500    (Al 6063-T5)

In other words, 3 times as great = so if a load at (say) 2 metres just bends the scaffold pole, then the same load could be extended to 6 metres and just
bend the pole.

*** Note that does not allow for the SLIGHT increase in wind load by extra length and diameter, but it is less than 10% when you do the Maths.

You can do the same sorts of calculations for the various diameters and wall thickness of other materials and compare. Interestingly, it is the outside
diameter, principally, that determines the strength of the system and that increases as the fourth power.

So a 20% increase in outside diameter = a factor of 1.2, becomes a factor of 2.07, because 1.2^4 = 2.07

PART 3: With the wall thickness that you are contemplating, the amount of wear and tear caused by any bolts or grub screws etc. that you might use to
hold the sleeve, will be minimal.

Why not consider an Aluminium welder spot welding the sleeve to the pole in  a few places, sufficient to hold the expected weight.?

PART 4: Personally I would favour the full weight of the mast and antennas sitting vertically on the rotator and the thrust bearing only being used for weight support when you may be changing out the rotator. That idea is based on the fact that most rotators are designed to withstand a vertical load in excess of 200 kg (e.g. Yaesu 2800 300kg, Emotator 1200 400kg KenPro 800 200kgs) and the bearing /inertia/breaking system appears to behave best when preloaded, not unlike rear axle bearings in a car.

PART 5: If you consider steel as an alternative, remember that all other things being equal, the weight of steel is twice as great as Aluminium, (due to its greater density) for the same maximum bending moment.

PART 6: Your starting off point has to be what increase in surface area will the larger array be? 2 times? 3 times? Once again, all other things
being equal, the load on the pole will be directly related to the wind pressure on the surface area of the antenna system.

If you feel that your existing system JUST manages in a strong wind and you are going to double the surface area of the array, then clearly you will be looking at something like you are suggesting at 60 or 63 mm OD. But don't assume that ALL steel will be stronger than aluminium for a given size, as you can see from the yield strength figures in the above table. C350 is the steel to go for if, you are to consider that. and it is generally available in 2.3 and (sometimes) 2.9 mm wall thickness

PART 7: As an example of what can be done - I recently did the design work for a change on David VK3EW's tower. He was using a 2.6 metre extension out of the tower of 48.4 scaffold tube supporting a 402BA 2 el 40mx yagi. The antenna was to be changed to a Coman 2el 40mx yagi at 4.0 metre extension, which has a similar surface area..

We went with C350 Duragal 60.3 mm OD 2.3mm Wall Thickness.


    AL tube:

        48.4mm diameter that is 2.6 metres long and gives a yield strength of:    191,441 x 250 / 2.6 = 18,407,788

    Now the Fe tube:

        60.3mm diameter this now 4.0 metres long and gives a yield strength of:    226,457 x 350 / 4.0 = 19,814,988

In other words we were able extend the height of the antenna by an additional 1.4 metres (to improve separation from a TH6 right at the top of the tower section) and still improve the safety factor by 10%

Naturally the weight was greater, but that was not the prime consideration, given that the tower was a tiltover/windup one anyway.

Hope this helps!
Peter VK3QI

3rd Feb 2009, Mike, VK1OO replies:

Hi Guy,

Here is an excel spreadsheet I developed to address this sort of problem


3rd Feb 2009, Peter, VK3QI replies:


That's a great spreadsheet.

If Guy tries inputting his current arrangements and then tries his calculated new array area, he will get a quick answer to his question.
(Assuming his new array doesn't change the effective height of the antenna)

The Yield strengths quoted, I agree with (always slight variations), but from recent experience in Melbourne from the Aluminium suppliers,
almost all tubing available in the size range required, is extruded. Only the telescoping tubes 1.42mm and 0.9mm WT are drawn.

The 6351-T5 is similar to the 6063-T834 but at a slightly lower 36,000 psi.

Peter VK3QI

3rd Feb 2009, Mike, VK1OO replies:

Hi Peter,

Thanks. I can't quite remember when I put it together. Formulas came from ARRL books. I seem to recall that I had trouble getting good data on available tubing.


3rd Feb 2009, Clint, vk3csj replies:

About 10 months ago I posted a similar request about a replacement pole, there were a number of good replies but still it left me unsure whether to use a length of Scaffold or go back to a length of Gal/Steel...well about a month ago I finally opted for a 5 meter length of High Tensile Scaffold with a 4.47mm wall thickness 4.48mm Diameter, same as what Guy has been using.

I wanted to use thicker wall stuff the next size was 50mm which I would've purchased but do you think any place in Melbourne had it in stock, the only place on earth was some joint in Queensland and would take a week to get down here so I opted for the smaller diameter, found a slightly small diameter tube which I have slid inside to hopefully add some extra strength at the point where the pole goes through the tower and thrust bearing.

Thank you Peter for your detailed post has helped a lot, I would've loved to have seen your same post 10 months ago, I was especially interested in you comment about the vertical thrust bearing too, I understand the most "good" rotators are designed to take a fair amount of vertical weight, however I have still tightened the thrust bolts to support the pole rather than the Rotators Clamping bolts but now I'm re-thinking that, in fact I've often wondered about the legitimate use of a vertical thrust bearing on a average Amateur installation.

Oh and use of 50mm Diameter Tube is about as far as you can go with a Nally Tower anyway due to the hole diameter of the support section that makes up the very top of the tower, therefore use of 60 or 60.3mm tube just wouldn't fit...

Again thanks Guy for asking and thanks to Peter for the interesting reply.


Clint - VK3CSJ

P.S. I have reduced my wind loading area up on the pole as well just to be sure to be sure, I have replaced that blasted pole (mast) so many times now that I'm sick and tired of the whole procedure of removing and fitting back antennas and using a million cable ties to do up the cable run into the shack, I have seven antennas up there 5 of which are horizontal beams, it just has to do!

3rd Feb 2009, Peter vk3qi replies:

From Guy's post, you will realise that a number of Nally style towers have been modified to take larger diameter poles.
Typically the 2 inch piece of pipe is cut out and a circular (horizontal) plate with a larger hole is welded to the three (vertical) flats and then a thrust bearing and its casing is bolted to the circular plate.

Your idea of a snug fit piece within the scaffold tube is OK, but really only of real actual benefit if the fit is EXTREMELY snug, and that is difficult to do a couple of meters inside of the tube.

Let's say you put a 2.5 mm thick piece inside the scaffold tube. (of the same yield strength)

Then the numbers compare:

(24.2^4 -19.68^4) x 250 versus (24.2^4 - 17.23^4) x 250 = 47,860,250 63,710,670

Just the same, every bit always helps!

Interestingly, if the aluminium was SOLID at that point, the result would be 85,743,500, which is only about 80% overall increase!

Your PS suggests that you might be in the market for some Chrome moly masting (very expensive) - however I would look at it this way.

You have to have some point on the tower/mast that is going to be the failure point in that once in 10 year wind shear event (that's code for a mini-tornado). Would you rather have the mast bend or have the tower fold over between the top of the bottom and bottom of top section, with the possibility of serious property damage to you or your neighbours? I know on my Nally tower that it is designed to fail at the mast junction for that very reason - even if it meant having to hire a bucket or crane to retrieve the situation.


Peter VK3QI

4th Feb 2009, Guy vk2ku replies:

My thanks to all those who replied to my posting, both direct and on the reflector.
I will get back to everyone today or tomorrow, and probably post my own thoughts here too.

However I am a bit busy on other non-radio things today!


4th Feb 2009, Guy vk2ku replies:

This issue seems to be of wide interest, hence this global reply.
There are several separate issues, and it is easier to address them separately.

My new array is intended for 2m EME, and will thus be larger than the antennas of most VK VHFers.
There are 2 separate severe conditions to which the mast is subjected, both of which act to subject the mast to a large bending moment.
1. Horizontal wind loading in the vertical mast position.
2. The weight of the mast and array when the tilt-over tower is being tilted into a horizontal position, and before the mast can be supported in that position.

The second of these seems to me to be at least as serious an issue as the first. Perhaps they are comparable. The calculations are not hard.
I am now pretty much convinced that for large arrays the "trolley"
system of lowering an array down the side of the tower is the best way to go, in that it completely overcomes the second issue above.
Nevertheless I have a tilt-over Nally tower already in position, so I intend to run with this arrangement (until it fails).

It was implicit in my original posting, but it is worth emphasizing that when I obtained my second Nally tower, I took certain design steps to reduce both the above problems:

1. The tower has only a single section (the bigger bottom section).
Increased height does not help much with EME, except when the moon is close to the horizon, and it lengthens the coax runs needed.
Those who have seen my first Nally will know that I keep it part lowered all the time, so as to increase the overlap between the 2 sections (and hence its strength). And no, the winch cable does not carry the load of the upper section, but rather this rests on a piece of hardwood through the lattice of the lower section.

2.The rotator mounting plate is set 2.4m below the top bearing, rather than the usual 1m, thus reducing the sideways force on the rotator plates by more than a factor of 2 times.

3. The usual bearing at the top is increased to 64mm to accomodate a 2.5-inch mast, the maximum size for the Yaesu rotator (G2800).

4. The tilt-over winch and cable is heavier duty than the one normally supplied.

Both bending moments can be minimized by setting the main cross boom which supports all the antennas as close above the top thrust bearing as possible.

Having done all that already, the only thing I can change is the material of the mast section, and one of my 2 original questions was directed at exactly that issue, rather than a calculation of the actual bending moment to be expected.
I will come back to mast material in a second posting.

I also asked for advice about how to attach a sleeve/collet to the mast just above the thrust bearing, so that the vertical load of the array would be carried on the thrust bearing rather than the rotator.
It seemed to me that there was some danger of weakening the mast.
I received some suggestions for this, but basically there are not a lot of options:
1. Weld the collet to the mast
2. Bolt the collet to the mast
3. Hold the collet in position by screws which work either by friction on the mast, or perhaps by indentations in the mast.

I was perhaps thinking too narrowly here.
The main cross boom already needs to be attached firmly to the mast, using some sort of gusset plate and U-bolts.
A simpler solution seems to me to fit a tubing sleeve over the mast below this plate, so as to sit on the thrust bearing and thereby carry the load of the cross boom (and necessarily the mast attached to it).

More anon ...


4th Feb 2009, Peter vk3qi replies:

Hi there Guy,

I look forward to your second response to the list regarding your proposed tower arrangements.

Having had a chance to look at the picture of your current setup on your website I can see that some of information you have received may not be entirely appropriate.

Assuming that you are going to expand your 144 setup from the present 4 long boom yagis, but still maintain a similar stacking arrangement and hence a main horizontal boom just above the tower top, then the issue of bending moments on the mast are not really the main issue.

I would have thought that the rotational torque on the mast, due to a larger
144 array would be the primary issue.

In that case, increasing the diameter to 63 mm is definitely worth while.

Firstly, as it makes setting up and aligning the rotator much easier (as shimming will be unnecessary at its maximum design mast size) Secondly, the increased rotational torque on the pole will be easier to handle for the rotator brackets.

Have you done any estimates of the rotational torque that will be created by the new array and whether the 2800 will be able to handle it?

Have you considered one of those torque absorbers that can be put between the rotator and the mast?

Have you considered some sort of feathering clutch/brake arrangement to hold the array in strong winds (much like the system used on some windmills)?

Personally, I would punt for a steel mast, as it will allow you to weld any bits and pieces on to the mast either above the tower or even within the tower where the rotator connects - all without reducing the strength of the steel to any extent.

Whereas, when you weld Aluminium (with a TIG), you have to derate the yield strength of the Aluminium 6061 and 6063 by a significant amount within a certain distance of the weld.

see: http://www.hollaender.com/files/1/Tech_Data/handraildesign.pdf

Your idea for setting the cross boom down on the collet and have the collet sit snugly on the bearing makes good sense.

Most thrust bearing arrangements I have seen in amateur situations have relied on the grub screws to do all the weight and centering work, but your idea will work provided you can ensure that the load is evenly distributed around the entire top edge of the collet. You could even make the gusset plate integral with the collet with a bit of welding and make it a permanent fixture on the mast. The boom clamps would be the removable ones instead.

It looks like a great project - all success to you!


Peter VK3QI

5th Feb 2009, Guy vk2ku replies:

Mike VK1OO has posted a nice spreadsheet for calculating the bending stress in a mast from wind loading and dimensions.
Pity about the ancient units! Just divide by 145 for MPa.

It is obviously very useful to model a mast in this way, though for a large array like mine, there are significant uncertainties on some of the effective wind areas.

In my case, as mentioned before, there are two distinct situations to be modelled: wind loading of the vertical structure, and weight loading when the tower is tilted horizontal.
The weight loading should be easier to calculate if all the data are known.

I also commented that my existing arrangement bends significantly when tilted horizontal, to the extent that I would not feel comfortable with any more weight load for fear of irreversible deformation of the mast.
However the mast and array have stood for 10 years in severe weather (at times) without any mechanical failure.
This fact has led me to focus on the weight loading more than on the wind loading as the limiting factor.

A simpler approach for me now is therefore to compare the planned new array with the existing array, on the basis that the bending moment shall only exceed that in the existing situation to an extent allowed by the stronger mast.

There is approximately double the amount of aluminium in the new array, but the new cross boom is much heavier.
Say a factor of 5 in total weight of the array.
But the new array will be centred no more than 150mm above the thrust bearing, compared with 500mm at present.
The net effect is an increase in bending moment by a factor of 5/3.3=1.5 compared with now.

The formulae in Mike's spreadsheet can be greatly simplified by making some reasonable assumptions:
1. For a large array, the loading contribution of the mast is negligible compared with that from the array.
2. The wall thickness of the mast is much less than its radius.

In that case for a given bending moment (from wind or weight), the maximum stress in the mast material (at the top tower bearing) varies inversely with its diameter and with the area of metal in the cross section.
This seems intuitively reasonable.
Put another way, if the diameter is D and the wall thickness t, then stress is inversely proportional to D^2 * t.
The squared factor arises because doubling the mast diameter also doubles the area of metal in the cross-section.

So increasing my mast diameter from 48.4mm to 63.5mm AND increasing the wall thickness from 4.47mm to 6.35mm reduces the stress by a factor of 2.4.

Overall I seem to have reduced the stress by a factor of 2.4/1.5=1.6 compared with the present arrangement.
Certainly I should be no worse off.

All of which brings me back to my original question!
What material should I use for the mast?
The options seem to be:
1. Aluminium alloy 6061-T6 (yield stress 241 MPa) - scaffold poles

2. Aluminium alloy 2024-T3 (yield stress 270 MPa?) - used in aircraft

3. Chromoly alloy (Yield stress 410-660 MPa) - bicycle frames etc

4. Mild steel (yield strength 250 MPa)

5. Hi Tensile steel (yield strength 635-690 MPa)

Mild steel is not an option, being too ductile, no stronger than 6061 aluminium, and much heavier.
2024-T3 and Chromoly are likely to be unavailable in the required size, and prohibitively expensive.
6061-T3 IS available in the current Capral catalogue in 63.5mmx6.35mm.
Hi-tensile steel is the other option if the right size is available; it is 2.9 times denser than aluminium alloy, but 2.7 times stronger so the wall could be that much thinner.
Weight for weight, there seems little to choose between 6061-T6 and hi-tensile steel. So it all comes down to price and availability.

Sorry for the length of this dissertation - it just grew!
But some may find the approach and the conclusions useful in their own situations.

Comments are of course welcome.

VK-VHF mailing list