Case study

Making an IKEA RÅSKOG work as a clothes rack

I adapted a RÅSKOG clothes-rack modification originally shared by home_design_notes. When I could not source the suggested parts in the UK, I designed customised lower and upper adapters to make locally available aluminium tubes fit the existing RÅSKOG connections.

Role

Problem framing, measurement, adapter design, prototyping & testing

Tools

Digital caliper, ChatGPT, Bambu Studio, Bambu Lab A1 3D printer

Year

2026

Customised lower and upper adapters shown beside a modified IKEA RÅSKOG mobile clothes rack.
I designed customised lower and upper adapters to make aluminium tubes available in the UK fit the existing RÅSKOG connections, adapting a clothes-rack modification originally shared by home_design_notes. (AI-generated visual based on my project photos.)

Overview

The original method relied on specific sizes of wooden rods and metal tubes that I could not find in the UK.

I came across a RÅSKOG modification shared by home_design_notes on Instagram and thought it was an interesting way to give my trolley another use. The original method used four 22 mm hollow metal tubes, with 20 mm wooden rods acting as internal connectors.

I could not find the required wooden rods and metal tubes in the UK. The closest suitable option I found was 19 mm aluminium tubing, but it could not connect directly to the existing RÅSKOG parts.

How could I connect parts with different diameters securely?

The challenge

My approach was to design two types of adapters to join the upper and lower RÅSKOG connections to the aluminium tubes.

I started by taking apart the RÅSKOG. Then I measured the existing connection points and the aluminium tube with a digital caliper.

Disassembled IKEA RÅSKOG trolley showing the original parts and connection points.
The RÅSKOG disassembled to understand how the existing parts connect.
Collage showing the lower and upper RÅSKOG connection problems and a digital caliper measurement of an existing part.
The upper and lower RÅSKOG connections presented different fit problems, so I measured the existing parts before defining the adapter requirements.

What I did

1. Starting with the lower connection

I initially aimed to use a sandwich-style fit to make the connection more secure and stable.

I measured the lower connection, including the internal crosspiece that limited the available space, and used these measurements to define the requirements for the adapter.

I translated these requirements into instructions for ChatGPT to generate the 3D-printable model, then prepared it in Bambu Studio and printed it on my Bambu Lab A1.

The resulting adapter fitted the lower connection as intended.

Four black 3D-printed lower adapters arranged upright on a tabletop.
Four completed lower adapters after the fit was finalised.

2. Moving to the upper connection

With the lower adapter working as intended, I used the same approach for the upper connection.

I measured the existing RÅSKOG part and one of the aluminium tubes, then printed a small test cap to check the fit before producing the full upper adapter.

The test cap fitted as expected, so I used the same dimensions for the first upper adapter. That adapter also fitted the tube I had measured.

As all four tubes were sold as the same 19 mm product, I expected the same adapter dimensions to work across all four.

Small test cap used to check the upper fit, followed by the first upper adapter fitted to the measured aluminium tube.
Testing the fit with a small cap before producing the first upper adapter. The completed adapter then fitted the aluminium tube I had measured.

3. The “19 mm” tubes did not fit consistently

When I tested the adapter with the other tubes, the fit was inconsistent.

Some tubes were tighter than others. From the physical parts, I could see that the difference appeared to come mainly from variation in the surface coating rather than a completely different tube size.

Instead of measuring every tube in detail at this stage, I chose a quicker way to test the fit directly.

Upper adapter fitted to one aluminium tube during an early fit test.
The adapter fitted the first tube, but the same dimensions did not fit the others consistently.

4. Testing three sizes directly

Rather than repeatedly changing the full adapter, I printed three small test caps in slightly different sizes and compared how each one fitted the tubes.

This let me quickly identify which size was too loose, too tight or secure enough, without needing to rebuild the whole adapter each time.

I then used the best-performing size for the revised upper adapter.

Three differently sized test caps compared on the aluminium tubes to identify a reliable fit.
Three test caps helped me compare loose, tight and secure fits before revising the full adapter.

5. The RÅSKOG parts also varied

After finding a reliable fit for the aluminium tubes, I assumed the original RÅSKOG connection points would be consistent too.

I measured one RÅSKOG connection point and printed a small test cap to check the fit. As it worked, and I already had reliable dimensions for the aluminium tube side, I produced four upper adapters.

When I tested them across the trolley, the fit was inconsistent. One adapter felt loose, while another would not fit at all.

This challenged my assumption that the original RÅSKOG parts were identical. I then measured the inside and outside diameter of each RÅSKOG leg twice to check the variation.

The measurements confirmed that the original parts also varied slightly, so I revised the adapter again to accommodate the differences.

Four upper adapters produced after the first RÅSKOG fit test succeeded.
I produced four upper adapters after the first test cap fitted successfully.
Eight caliper readings showing outside and inside diameter measurements across four RÅSKOG legs.
Grid of caliper measurements showing the outside and inside diameters of four RÅSKOG legs.
Leg 1Leg 2Leg 3Leg 4
OD (mm)1st: 19.43
2nd: 19.44
1st: 19.36
2nd: 19.41
1st: 19.30
2nd: 19.65
1st: 19.44
2nd: 19.53
ID (mm)1st: 15.27
2nd: 15.39
1st: 15.66
2nd: 15.65
1st: 15.42
2nd: 15.70
1st: 15.76
2nd: 15.84
OD = outside diameter; ID = inside diameter

6. One RÅSKOG connection still did not fit

Even after adjusting the adapter for the variation between the RÅSKOG legs, one connection still would not fit.

I inspected that position more closely and found a small burr around the opening that reduced the available clearance.

This showed that the remaining problem was local to that connection, so I needed a targeted adjustment rather than another change to the whole adapter.

Close-up comparison of two RÅSKOG leg openings, with a small metal burr visible on one opening.
A small burr at one RÅSKOG connection reduced the available clearance.

7. Making a targeted adjustment

Instead of making the whole adapter smaller, I changed only the area that was causing the interference.

I added a wider clearance slot to the adapter so it could pass the burr at that specific connection while keeping the rest of the fit unchanged.

Rather than measuring the exact distance between the tube edge and the burr, I used the physical interference itself to guide the adjustment. This reduced the number of measurements I relied on and avoided introducing unnecessary error.

I printed the revised version and tested it again. The updated adapter fitted the previously blocked connection while preserving the overall fit elsewhere.

Revised 3D-printed adapter with a wider local clearance slot to avoid the burr.
The revised adapter added clearance only where the interference occurred.

8. Finalising the adapter designs

A collection of 3D-printed adapter prototypes and fit-test pieces produced during the project.
Across the project, I used multiple fit-test pieces and adapter iterations to refine the final dimensions.
Final upper and lower 3D-printed adapters arranged in two groups on the printer bed.
Finalised adapter designs after physical prototyping: upper adapters (left) and lower adapters (right).

9. Final assembly and real-world use

Once the adapters fitted reliably, I assembled the full trolley and tested it as a working clothes rack.

I installed the four aluminium extensions, rebuilt the RÅSKOG and added the top rail for hanging clothes. I then tested the completed trolley with clothing and used it in its intended space. The adapters held the structure securely and the modification worked as a complete system, not just as individual fitted parts.

Completed RÅSKOG clothes rack shown after assembly and in use with clothing hanging from the top rail.
The completed trolley after assembly (left) and during testing with clothing (right).

Outcome

The adapters made it possible to recreate the clothes-rack modification using parts I could source locally.

The final setup kept the original trolley while adding the hanging height I needed. The 3D-printed adapters allowed the locally available aluminium tubes to work with the existing RÅSKOG structure, and the completed rack has remained stable in everyday use.

Completed modified RÅSKOG trolley used as a clothes rack in its intended space.
The completed RÅSKOG modification in everyday use.

Reflection

This project reminded me that one successful test is not enough to prove that a solution will work in every case.

It also showed me that nominal dimensions and product specifications should not be treated as perfectly exact. Even parts that appear standardised can vary slightly in the real world, and those small differences can matter when the fit needs to be precise.

The four lower adapters fitted as expected, and the first upper connection appeared to confirm that the same approach would work again. Testing the remaining parts showed otherwise: both the replacement aluminium tubes and even the original RÅSKOG connections varied enough to affect the fit.

Each time new evidence challenged my expectations, I went back to the physical parts, isolated the cause, made a targeted change and tested again.

This helped me avoid changing more than necessary and gave me more confidence in each decision.

Although this was a small personal project, the same principle applies to digital services: a solution that works for one case or one user should not automatically be assumed to work for everyone.