Teardrops n Tiny Travel Trailers

My wife and I fell in love with the elegance of design represented by Mr. McCamant's Travelear. So we are building one for us that we intend to pull behind our Prius. To do this we are working to get the total weight in below ~750#. We are using an aluminum frame that we made from scratch. To keep the weight low we opted not to weld but use structural adhesive and rivets. Attached are some pics of the progress.

I like it.
I did not stay at a Holiday inn last night nor am I an engineer.
I am questioning the tail end of the a frame under the perimeter of the frame..
Others will chime in but that makes me wonder..

Thanks for the question. If the A-Frame was integrated on the same plane as the main chassis the points of penetration in the chassis would require complex brackets to keep the original strength. Ultimately the frame would be weaker. By having the A-Frame under the main chassis the A-Frame triangulates the forward corners and only adds strength. It also allows the addition of 'doublers' at the point where the chassis lays on the A-Frame. Below is a photo of the doublers I added at the Axle points. Doublers are an old aircraft fabrication trick often used in aluminum.

Lookin' good so far... Are you planning commercial doors?

We did consider that. But the look of the matching aluminum door is just too compelling. As terrifying as it seems we are going to try to match the scratch built doors on the Travelear.

Quick update we made some progress on the skinning of the roof. We were able to get the wiring roughed in and get part of the roofed skinned.

You certainly have chosen a unique way of building your chassis.
Some questions, if you don't mind:
1. What size & type of channel?
2. What size & type of rivets?
3. Length of tongue?
4. Size of 'doubler' angles?
5. How did you determine the tongue strength? We have a formula for steel tongue design, but I have yet to see a formula for aluminum.
6. A close up picture of how the tongue rails are attached?
7. The weight of the chassis w/o axle?
8. And finally, why did you not run the tongue rails farther back on the side rails?
These are all serious questions, because I'm very interested in what you've built.
Thank you,
Fred

What a great and meaty set of questions, very cool!

1. What size & type of channel? 6061-T6 3"X0.170" structural channel American Standard Channel not the Architectural Aluminum channel profiles
An important point here would be to stay away from square tubing. When you fasten anything to an aluminum tubing you setup crush modes.
For a trailer that uses aluminum tubing when you fasten the axle it will tend to crush the tube. As soon as this starts you can never tighten the bolts enough have it work.
When you use channel the bolts can mate to solid faces of aluminum-tighten all you want.

2. What size & type of rivets? They are all 1/4". Most are 3/4" long. A few of the doubler rivets are 1/2" long.
One of the things you have be careful with aluminum is corrosion. As long as there is not another metal it will be very corrosion resistant.
If you use steel rivets the aluminum will corrode quickly. This is why you will see me add duct tape to the steel jacks and axle where it joins the frame.

3. Length of tongue? It is based on the Atma Travelear design so it come out to ~50"

4. Size of 'doubler' angles?
The doubler angles are 1/8" rectangular 6063 Architectural Aluminum. I used the rectangle so that when I cut it the edge would mate with a tight glue line to the main channel.
The size is set to mate with the entire frame surface and be 50% longer than the contact point for the axle and 2X at the tongue lever point.
All doublers are riveted and glued. The glue is two part SEM structure adhesive. Most of the acrylic formulas work great on aluminum

5. How did you determine the tongue strength? We have a formula for steel tongue design, but I have yet to see a formula for aluminum.
This is tricky to explain in text. In the Atma Travelear design David has the lower side build with a relief-dado. I use that and the skeleton-ply layers of the sides to serve as stressed members.
The wood sides are glued and screwed to the aluminum rails with the axle on with tires. The frame is suspended to stress the center. Then the sides are mounted. This allows the sides to stiffen the main frame rails. There is another longitudinal center aluminum rail and the torsion axle is a stressed member. All that is stated to explain that the trailer is a separate stressed element. This allows the tongue to be evaluated as a simple lever. I will try to setup a quick spreadsheet for the PSI/length calculations.

6. A close up picture of how the tongue rails are attached?
Will do on the photo. May take a few days-cold front hit us in Texas and it is cold out there-I am sure the folks from AK are amused...
In short there is a doubler at the front edge lever point that extends ~14" on each rail. There is also another cross brace that is not yet added because of some tongue box fab that I am waiting on.
The actual attachment is with 5/8" grade 8 bolts, washers, etc...

7. The weight of the chassis w/o axle? -98# total

8. And finally, why did you not run the tongue rails farther back on the side rails?
Not wanting to weld I could not match David's design for a single strut that extended to the hitch. So I used a 50 degree hitch.
Match 50 degrees to the length of the design from David and you ~50". Would it be better if they were further back-yes. But really would be overkill.
I am a pretty big boy and I can jump on this thing-it is stiff and solid.

I hope this helps... I will get to work on an updated attachment photo and simple spreadsheet.

These are all serious questions, because I'm very interested in what you've built.
Thank you,
Fred

Thanks for those explanations.

All that is stated to explain that the trailer is a separate stressed element. This allows the tongue to be evaluated as a simple lever. I will try to setup a quick spreadsheet for the PSI/length calculations.

Yes, please do.

In short there is a doubler at the front edge lever point that extends ~14" on each rail. There is also another cross brace that is not yet added because of some tongue box fab that I am waiting on.

Will the tongue box be an integral firmly attached part of the cabin so that 'lever point' is farther forward? i.e. the tongue is actually shorter than 50".
Fred

Well Fred you are a powerful thinker. I will try to explain in a manner that is as helpful as possible-meaning I am sure others may be able to explain this better than me so please jump in.
Also please take this information in the context it is given, meaning don't call me if something breaks...

In what I am about to share there will be one term that may seem odd it is denoted as "I" or the "moment of area". It is a number that is represented by the cross section of a profile (beam, channel, rod, etc..) and a factor related to the profile shape. In simple terms the value "I" reflects a given profile's ability to resist bending. I beam will be resist bending in one direction better that another...

So to figure out how much a beam like a tongue will deflect we need to know the value of "I" for our structural profile. Go to this website (this one is for aluminum, there are others out there for steel) and click on one of the shapes, enter the shape dimensions and it will give you a number for "I" (Ixx). Get that number and then...
http://www.amesweb.info/SectionalPropertiesTabs/SectionalProperties.aspx

Grab this spreadsheet. It is made for artists like us so it is simple.
http://wwwedu.artcenter.edu/mertzel/structint/structures%20worksheet.xls
Enter the values for a cantilever beam and the deflection you can accept then read the value at E18 (concentrated load).
If the value for (Ixx) of your planned material is greater or close to the profile value you are good to go. If not, get bigger material.
The spreadsheet also works in reverse to tell you what (Ixx) you will need.


"Will the tongue box be an integral firmly attached part of the cabin so that 'lever point' is farther forward? i.e. the tongue is actually shorter than 50".
Fred"
That is a very interesting thought. I feel pretty silly that I had not considered this before. So my answer is YES. Based on this suggestion (in the form of a question) I will be locating points on the cabin to triangulate the tongue box into. This is a very clever thought and I am grateful you took the time to ask. And yes that will effectively shorten the tongue length and further stiffen the structure.
Thanks greatly Fred!

One last though-my full photo log is here.
"https://plus.google.com/photos/107496989082639774873/albums/6061482475601687185"
There may be some photos here that will make things clearer.
We expect warmer/wetter weather this weekend, so I should have some updates by Monday.

Don't know about being a powerful thinker, Bryant.
Just an old guy without a formal education that likes to design & build things. My only practical experience with using stressed aluminum was designing & fabricating three 12 ton capacity welded aluminum fishing boats and net hauling machinery. They floated & pulled 3-4 tons/hour.

I have considered using aluminum for a trailer chassis, but have yet to find a satisfactory formula that I could trust without overbuilding & weighing as much as steel. So please take my questions & statements in that vein.

Rather than a cantilevered beam, wouldn't a trailer tongue be considered a simply supported beam with a concentrated load? The ends being the axle & hitch. The load being concentrated at the front of the cabin.

Also, isn't the spread sheet that you provided for a static load rather than a dynamic load? Meaning that the load is constantly changing as the trailer is towed. This is the big question for me when considering aluminum, because it cycles way fewer times than steel before it cracks & breaks.

Some time ago member Paul C from Australia provided us with the Australian trailer laws which include tongue strength rules. Member Angib (Andrew) from England interpreted them & published them in the Design Library (click Design Resources at top of the page).
http://www.angib.pwp.blueyonder.co.uk/t ... tear84.htm

The formula relies on vertical section modulus (in3), yield stress (lb/in2), tongue length, and tongue design (single, double, triple). Example: 36" 'V' tongue using 1/8x2x1 steel rectangular tube.
.280 (in3) x 31,500 (yield stress for closed section (lb/in2) = 8,820 x 2 (two rails) = 17,640 ÷ 18" (1/2 the length of the rails) = 980# tongue rating.

The yield stress for an open steel section (channel like yours) is 26,250 psi. For aluminum you could cut that by 1/3 (per your structure spread sheet) edit - should have said - divide by 3 and see where you're at. BUT that still doesn't address the reduced number of cycles that aluminum will withstand. Any ideas?

Fred

I believe the thrust of your questions cover two points:

a: Cantilevered beam vs simply supported beam with a concentrated load
b: Stress cycle resistance before cracking

For the sake of all that may read this I may cover some old ground, I know from the work I have seen on your trailer you already know most of this.

For question a: I believe either method can be used. The reason I selected the cantilever is that all bend stresses are comprehended in reference to that single element.
If I used the beam method I would need to address the actual joint itself. That is probably a better approach, but I tend to default to simpler methods for complex problems.

For question b: Yes I had the same concern. The big issue with trailers and other road structures cracking is the use of welding. Take a look at this video from the 1 minute mark on.
https://www.youtube.com/watch?v=7_q8dXhewh4
What you see is the cracks all form in the HAZ (heat affected zone) and are the result of welding. This is why so many Aluminum trailer manufactures are moving to VHB tapes and other products even over rivets. As for the problem of cycle stress fractures that can occur even if welding is avoided-I would make two points. First much of how this is avoided is the result of heat treating and artificial aging. It may even be fair to state that the T6 as a treatment is the result of the needs of aircraft to resit many cycles without cracking.

For most aircraft the design limits are somewhere between 25,000 and 80,000 cycles. In fact there are commercial aircraft for sale today that have more that 75,000 cycles. I would suggest those numbers are beyond any stress that recreational vehicles would endure. But there is the issue of dynamic loads and you are correct the spreadsheet models only statics. So a simple method would be to take the maximum G-loads from road stress and design for that level. Except that a typical pothole exposes most vehicles to acceleration of 50-80Gs. What happens in real life is that those stresses occur through the tires/axles/suspension and are filtered/dampened before they get to the other structures. In addition the time period of the impact, although a very high in load is very short in time. All of these things tend to reduce the actual impact to the rest of the structure. A good example is a rifle. Consider that when you fire a large caliber hunting rifle the force that is applied to the shooter and the force at the muzzle are the same. The only reason the shooter is not harmed is that the force is spread through 8-10 sq inches of contact area and the force is dampened by the mass of the rifle itself. If the rifle is a semi-auto, the bolt and gas/springs further dampen this load over time. So a load that is clearly lethal becomes very safe.

I believe though what you are getting at is this (is there a simple way to KNOW a particular structure is light and safe for all/most conditions?). Yes there is, but it requires higher order math and is not something that can be implemented easily in a spreadsheet. I am sure is it possible to integrate some dynamics in spreadsheets. I have seen some that have used fourier transforms and mohr's circle in spreadsheets. But most would just use some FEA(finite element analysis) SW. If you have an interest in this area take a look at Lisa or VisualFEA. Both are free for limited structures.
http://lisafea.com/
http://www.visualfea.com/download.htm

In reference to the practical design elements of strong design take a look at the "Handbook for Damage Tolerant Design". This is a great resource you may find helpful.
http://www.afgrow.net/applications/dtdhandbook/sections/page9_2_1.aspx

Thank you, Bryant, for all this additional information. Very helpful.

Debraizhot wrote:I believe the thrust of your questions cover two points:

a: Cantilevered beam vs simply supported beam with a concentrated load
b: Stress cycle resistance before cracking

For the sake of all that may read this I may cover some old ground, I know from the work I have seen on your trailer you already know most of this.

For question a: I believe either method can be used. The reason I selected the cantilever is that all bend stresses are comprehended in reference to that single element.
If I used the beam method I would need to address the actual joint itself. That is probably a better approach, but I tend to default to simpler methods for complex problems.
After looking at the Lisa program, I see your reasoning.
For question b: Yes I had the same concern. The big issue with trailers and other road structures cracking is the use of welding. Take a look at this video from the 1 minute mark on.
https://www.youtube.com/watch?v=7_q8dXhewh4
Wow, you wonder what he hauled on that trailer. I did note that he never mentioned the tongue though.

What you see is the cracks all form in the HAZ (heat affected zone) and are the result of welding. This is why so many Aluminum trailer manufactures are moving to VHB tapes and other products even over rivets. As for the problem of cycle stress fractures that can occur even if welding is avoided-I would make two points. First much of how this is avoided is the result of heat treating and artificial aging. It may even be fair to state that the T6 as a treatment is the result of the needs of aircraft to resit many cycles without cracking.
I don't weld across the top of the steel rails for that reason & can see why, with aluminum, using rivets, bolts or clamps, as you have, would be a better choice than any welding at all.

For most aircraft the design limits are somewhere between 25,000 and 80,000 cycles. In fact there are commercial aircraft for sale today that have more that 75,000 cycles. I would suggest those numbers are beyond any stress that recreational vehicles would endure. But there is the issue of dynamic loads and you are correct the spreadsheet models only statics. So a simple method would be to take the maximum G-loads from road stress and design for that level. Except that a typical pothole exposes most vehicles to acceleration of 50-80Gs. What happens in real life is that those stresses occur through the tires/axles/suspension and are filtered/dampened before they get to the other structures. In addition the time period of the impact, although a very high in load is very short in time. All of these things tend to reduce the actual impact to the rest of the structure. A good example is a rifle. Consider that when you fire a large caliber hunting rifle the force that is applied to the shooter and the force at the muzzle are the same. The only reason the shooter is not harmed is that the force is spread through 8-10 sq inches of contact area and the force is dampened by the mass of the rifle itself. If the rifle is a semi-auto, the bolt and gas/springs further dampen this load over time. So a load that is clearly lethal becomes very safe.
Still a conundrum for me. After reading the DTD handbook page, however, it occurred to me the it would be interesting & useful to attach sensors to the tongue in order to measure G-forces, deflection & cycles/time graphed out over different road conditions.

I believe though what you are getting at is this (is there a simple way to KNOW a particular structure is light and safe for all/most conditions?). Yes there is, but it requires higher order math and is not something that can be implemented easily in a spreadsheet. I am sure is it possible to integrate some dynamics in spreadsheets. I have seen some that have used fourier transforms and mohr's circle in spreadsheets. But most would just use some FEA(finite element analysis) SW. If you have an interest in this area take a look at Lisa or VisualFEA. Both are free for limited structures.
http://lisafea.com/
http://www.visualfea.com/download.htm Over my head!

In reference to the practical design elements of strong design take a look at the "Handbook for Damage Tolerant Design". This is a great resource you may find helpful.
http://www.afgrow.net/applications/dtdhandbook/sections/page9_2_1.aspx

Out of curiosity I did try some numbers with the Australian formula. 1.10 in3 x 8,750 lb/in2 = 9,625 x 2 = 19,250 ÷ 16 = 1,203# weight rating. (that was assuming an 18" box)

Anyway, it looks like you're building a real fine teardrop. And I'm probably out of my league conversing with you! If I may ask, what do you work at? Aerospace engineer? Physics teacher?

Thanks again, Fred

Thanks for the comments Fred. I hope there was some food for thought that was helpful. I truly enjoy the dialog. I did some design work when I was at General Dynamics, back then I worked in flight simulation. A good example of the kind of work back in the day: http://www.google.com/patents/US4687054
Today-I work for Airbus. My current job is Sourcing Manager-mostly contracts and agreements. http://www.linkedin.com/in/bryantunderwood/
Feel free to contact me any time. Trailers are a fun relief, plus it makes Ms. Debra happy.
Also with regard to 'not in my league conversing with you!' believe me-that is not possible.