Lightweight sandwich panels

Here we will calculate the theoretical minimum weight of sandwich panels. This is used for my solar bike, the Maxun One. The sandwich panels that I need are extremely lightweight, as far as I know they are not available anywhere, I have contacted almost all factories. Therefore I wanted to find out if it was possible to build these myself.

Ultralight sandwich panels
Ultralight sandwich panels

The solar panel size is 900 x 650 x 10mm. For the first solar bike, I used Alustep 300 Light sandwich panels. These have glass fiber skins and an aluminum honeycomb core, the weight is 1,6kg/m2. A lot of experience has been gained with these panels, since they have been used more than 7000km. The stiffness is not high enough. Many solar cells on my solar bike are cracked because the panels bend too much during extreme fibrations. By the way, these cracks are no problem with the Sunpower solar cells because the wiring at the bottom of the cells still ensures good electrical connections. I want to have much lighter panels for the next solar bikes because a low weight will burden the panel holder construction much less on bad roads.

The goal is panels with better stiffnes than Alustep 300 Light and a weight of about 1kg/m2 instead of 1.6kg/m2. 

The panel strength is less important because the panels are not exposed to large external forces. The main forces are caused by their own weight when they vibrate while cycling on bad roads.

Lightweight sandwich test panels
Lightweight sandwich test panels

The calculation of the theoretical minimum weight of sandwich panels

The most lightweight skin fabric

For the skins I will use TeXtreme carbon fiber. TeXtreme uses spread tow fabrics which increases the performance as well as reducing the laminate weight up to 20%:

TeXtreme carbon fiber
TeXtreme carbon fiber

Step 1 Determination of the required skin thickness

To find out the required skin thickness, I have built and tested several sandwich panels. The bending strength is measured with a simple four-point bending test:

Four point bending test
Four point bending test

The panels are built up with TeXtreme carbon fabric and Aramid (Nomex) honeycomb. Because of the wind drag, the maximum thickness is 10mm.

It turned out that the sandwich panel has same strength as the Alustep if it is built up as follow:

  • Core: Aramid honeycomb Cormaster C2 C2-3,2-29 290g/m2, thickness 10mm, cell size 3,2mm
  • Skin: TeXtreme 80g/m2, laminate weight 180g/m2, thickness 0,2mm. I made this with vacuum bagging (see below).
  • Skin to core bonding: an abundant amount of 100g/m2 epoxy has been used per skin since weight plays no role in this test.

The weight of this test panel is 850g/m2.

Step 2 Calculation of the minimum panel weight without the skin to core bonding

The test panel (see step 1) was manufactured just to determine the strength, but no attention was paid to low weight. The weight of the skins can be reduced by using resin infusion instead of vacuum bagging. The laminate weight then becomes 130g/m2, calculated with this online tool:

So, without the skin to core bonding, the minimum weight up to now is 550g/m2:

  • Core: Aramid honeycomb 290g/m2
  • Skins: TeXtreme 130g/m2

Step 3 Calculation of the minimum skin to core (adhesive fillet) bonding weight

The weight of the skin to core bonding plays a serious role in the total panel weight. Therefore only glue has to be applied between the honeycomb cell walls and the skin in the form of adhesive fillets. Note that epoxy is used for laminating but it is not the best glue for skin to core bonding. It is better to use glue that has been developed specially for this purpose. 

Now I will draw up a formula for the skin to core bonding weight in relation to the fillet radius and the honeycomb cell size:

bonding weight / m² = density * adhesive fillet area * fillet length per cell * number of cells

cell size = s
adhesive fillet radius = r
glue density = σ

Honeycomb core cell geometry
Honeycomb core cell geometry


Adhesive fillet area
Adhesive fillet area

adhesive fillet area (if the fillet shape is circular) = r² -  1/4 π r² = r² (1 - π/4) 
fillet length per cell = 6L = 6 * s / √3
number of cells / m² = 1 / cellarea = 1 / (1,5L²√3) = 2 / (s² √3)
bonding weight = σ r² (1 - π/4) 6 * s / √3 * 2 / (s² √3) = σ r² (1 - π/4) 12 * s / (√3 s² √3) = σ r² (1 - π/4) 4 / s 

The bonding weight = 0,858 r² σ / s

This is the bonding weight for a honeycomb with cell size (s) = 3,2mm and with epoxy bonding (σ = 1150kg/m³) in relation to the fillet radius:

adhesive fillet radius [mm] bonding weight [g/m²]
0,3 28
0,4 49
0,5 77
0,6 111

Sandwich panel adhesive fillet
Sandwich panel adhesive fillet

Now we must determine the required fillet radius. According to this study, the optimal bonding weight can be 40g/m2. This corresponds to a fillet radius of 0,35mm in our example. But I doubt the strength of the skin to core bonding at this low weight. It is difficult to achieve a strong skin to core bonding; so it is unwise to save too much weight in this area. To be on the safe side, take a bonding weight of 100g/m2.

How strong should the skin to core bonding be? A rule of thumb is that it is strong enough if you have difficulty peeling it off by hand.

I did a test with dipping the Nomex core in epoxy until the weight was exactly 100g/m2. The bonding was still bad, even with this amount of glue, the fillets were not shaped:

Sandwich panel skin to core bonding example 100g per m2
Sandwich panel skin to core bonding example 100g per m2

The final weight of the sandwich panel

The final panel weight is:
core + 2(skin + bonding) = 290 + 2(130 +100) = 750g/m².

This is still a theoretical minimum value. I do not think this can be realized in practice. I'm happy with a weight of 1kg/m2, which is quite an improvement.

Rohacell 31 test

I have tested core material from Rohacell 31kg/m3. However, it is too easily compressible by hand. This is unsuitable because the solar cells are extremely vulnerable.

Sandwich panel skin to core bonding with heat activated adhesive film

To save weight, it is necessary to use a method where we can apply exactly the minimum required amount of glue in the form of fillets concentrated around the edges of the honeycomb. One method we could use is dipping the honeycomb core in glue, but this will soon become a big mess. It is more convenient to use a heat-activated adhesive film, which is dry during processing. Due to weight savings, we cannot use the adhesive film in the normal way. We only want to use the glue and not the film itself. This can be achieved with the following method. This is a so-called multi-step procedure, first one face will be glued and then the other.

The film is applied to the top of the honeycomb, pressed and heated in an oven until the glue becomes soft and sticky. Then the film is removed and a certain amount of glue will remain around the edges of the honeycomb. This must be done quickly to prevent the glue from hardening. When the glue has cooled down and is dry, the skin is applied to the core, pressed and then glued in an oven. The amount of glue must be precisely dosed by setting the time in the oven and the temperature.

Julien Rion has done a great research on ultra-light sandwich panels: Ultra-Light Photovoltaic Composite Sandwich Structures.

Electrical insulation

For the PV panel, the sandwich panel has to be additionally insulated on both sides because of the solar cells and the wiring. For this, I use the thinnest glass fabric of 25g/m². This gives a weight gain from 2 x 36g/m2 = 72g.


  • A superfluous adhesive layer of 0,1mm thickness causes an extra weight of about 110g/m2.
  • The core-skin bonding is very crucial for the final sandwich strength and it is quite a challenge to apply just enough glue for getting a nice adhesive fillet. 

I'm currently looking for someone that can make strong sandwich panels of about 1kg/m2 as described above.
Who can help?


Extensive studies have been carried out by other people on this subject, the following links are particularly interesting:

Heat activated adhesive films

A simple DIY sandwich panel four-point bending test

A four-point bending test is more convenient to use than a three-point test because the weight has automatically two supports points. The alustep 300 light panel serves as reference because this panel is known. Its bending is 2mm +/- 0,1mm with a load of 13kg in this setup:

Sandwich panel four-point bending test
Sandwich panel four-point bending test

The used vacuum bagging method

  • Vacuum bag
  • Breather
  • Impervious release film
  • Bleeder
  • Perforated release film
  • Peel play
  • TeXtreme fabric with 140g/m2 epoxy
  • PVA
  • Glass panel
  • Vacuum bag

Do you have any comments about the website? Please let me know.