Engineering Analysis: Grease or Retaining Compound

Posted on 14 Comments
Loctite 641 Retaining Compound


Littered in the comments section of almost every Hambini video is a comment about using grease, not using grease or the use of retaining compound. There is lots of misinformation on the internet that has been distributed by the mainstream bicycle media, Forums and others. This post will cover the merits of grease and retaining compound from an engineering perspective.

What is Retaining Compound?

Retaining compound is often generically called “Loctite”, especially in Cycling Circles. This is an incorrect term as the parent company is Henkel and Loctite is the brand and has many different types of retaining compound.

Retaining Compound is essentially a type of anaerobic engineering adhesive that is designed to take up imperfections between cylindrical fitted components. This is why it is called “bearing retainer” in engineering. The exact chemical method of activation will not be covered here. In practice, a small film is applied between the components, the absence of air initiates the action and some 72h later (more on this further down), a bond is created between the cylindrical fits. It is extremely useful where a system assembly needs to be disassembled at some point in the future as the mating surfaces suffer minimal abrasion from installation and removal.

Retaining Compound is available from multiple manufacturers and has different properties. The most common brand is Loctite and the two widely used variants are Loctite 641 and 638.

Loctite 641

Loctite 641 is a medium strength retaining compound. This is by far the most widely used type of retaining compound throughout both cycling and the wider engineering industry. This achieves near full strength within 72h (activator is often required for this speed of cure). It is suitable for metal surfaces and works with plated surfaces and carbon fiber with the use of activator. The manufacturer recommends a maximum gap of 0.1mm for use with this product.

Loctite 641 Retaining Compound
Loctite 641 Retaining Compound

Loctite 638

Loctite 638 is a high strength retaining compound. This has similar properties to Loctite 641 with the exception being it’s ability to work effectively in much larger gaps. The manufacturer recommends a maximum gap of 0.25mm for this product. In practice, carbon fibre and plated surfaces require activator to attain sufficient bond strength.

What is Grease?

Grease is a viscous petroleum mix that is designed to lubricate moving parts or aid assembly. It comes in different grades and chemical compositions and it has been covered in detail the the: Bearing Grease a Practical Guide post.

For general applications, an NGLI 2 grease is used.

The Mechanics of Installation

Instances where the debate between retaining compound, dry fit and grease come up usually involve fits of some description. In engineering terms, a press fit is an interference fit where by mating surfaces of components have sizes such that a mechanical force is required to press one item into the other.

In a perfect interference fit both components are concentric circles. The mating process causes one component to be in compression (the male component) and the other to be in tension (the female component). The radial stresses are sufficient to overcome the maximum axial force and the component stays fixed. The fit is perfect so there is an even distribution of radial forces across the componentry

In real world applications there is some level of radial or axial irregularity which means there is non uniform distribution of stress between the parts. In these scenarios, Retaining compound “takes up the slack” and produces a stronger retention force.

One of the biggest draws in recent years has been a drive towards sustainability and reducing overall life cycle costs. The tribology industry has responded and the fit tolerance required has been consistently reducing over the years.

Modern Engineering Methodology – the case for J fit

For many years, bearing manufacturers recommended housing or shaft fits in the range of m7 or M7. In recent times they have moved towards having a J or JS fit (depending on load). This is on the boundary between interference and transition. The fit is combined with the use of retaining compound for installation. The process of installation takes longer – it takes 72 hours for the retaining compound to achieve reasonable shear strength but the advantage comes during disassembly. Moderate heat (hair dryer) reduces the shear strength of the retaining compound and facilitates easy disassembly and replacement. There is much less wear on the expensive component which is typically the shaft or the housing. J fits often negate compressive or tensile tolerance stack ups which is an additional advantage.

There are some disadvantages with this methodology. The cure time has already been mentioned. Another key issue which is prevalent in the bike industry is misalignment and ovality. The apertures need to be round and aligned well (bottom brackets). A prime example of the problems associated with this is Trek’s BB90 bottom bracket standard. Trek were one of the early adopters of the shift for sustainability but their BB standard proved to be problematic. Their execution did not place enough emphasis on the need for a reasonably tight hole tolerance. They even went so far as to spec 0.1mm oversized bearings at one point.

It’s noteworthy that in the vast majority of cases, a set of normal bearings at the top end of the size spectrum (as used in the Hambini kit) was sufficient to overcome the problems. In pure engineering terms this is simply a tightening of the standard from J7 to J5.

Another key but often overlooked issue is the perception of the customer. Customers have long expected interference fits between the components so the concept of having what appears to them as a slack fit is associated with poor manufacturing. It will take some time for this engrained idea to be overwritten.

Dealing with Bike Manufacturer’s notorious Tolerances

Bike manufacturer’s have long been roasted on the Hambini Barbecue for supplying frames that do not meet their stated tolerances, they have never openly admitted to this and the nearest that any supplier came to an idea of admission is when OPEN cycle famously said “There was no documented standard they were required to meet”, this was in the midst of trying to sue the Hambini YouTube Channel.

Retaining compound and the use of a transition bearing fit does give more wiggle room to deal with undersized bottom bracket holes which are now commonplace in the market. It is not good engineering practice to design for someone else’s inability to manufacture correctly but as a workable solution, it is valid.

Grease attack

There have been reports primarily from the aerospace industry of grease attacking carbon fiber. Strictly speaking the mechanism of attack is to to soften the resin binder material within the composite – it does not attack the carbon itself. Modern carbon fiber composite appears to be fairly immune to this type of attack.

Early carbon frames are still susceptible so it is worth checking.

Of more concern is the frequent specification of poor grade seals in bicycle applications. Cheap bearings, often those found on Amazon and eBay, are fitted with seals that are reportedly nitrile. Genuine Nitrile is resistant to most greases (Shell Gadus, Mobil etc) but there have been reports and pictures of grease attack. In these scenarios it’s almost always due to low grade and misrepresented materials in the bearings.

Grease vs Retaining Compound

Despite the feelings of many armchair bike mechanics, there is no real reason to use grease for static applications. Using grease as an assembly aid is perfectly acceptable but using grease to silence creaking in applications between static components would better be solved with retaining compound. Creaking is movement that omits an audible frequency. The aim is to stop the movement.

Grease is certainly a valuable addition when fits are between metallic materials that are prone to galling – eg aluminum or stainless steel. As an example, using grease between an aluminum bottom bracket and aluminum bike frame would be recommended. However using grease between a metal bottom bracket and a carbon bike frame where all sizes were within permissible range would not be recommended – the grease would reduce the retention force.


The take away from this is the product needs to be correct for the application and that is dependent on what it is being used for and the level of fits between the components. Light fits on static components almost always need retaining compound. Heavy fits between metallic components would be better suited to grease – this is to reduce the chance of galling and installation ease, it is not for performance improvements once installed.

If you are unsure, then the recommendation would be to use retaining compound. It will act as a lubricant for installation (admittedly it is not as good as grease) and “fill” the gaps once it has set.

14 thoughts on “Engineering Analysis: Grease or Retaining Compound

  1. That’s a relief. I was worried at first because I just installed new SKF bearings in my hubs and I used NGLI 2 grease instead of retaining compound.

  2. The bike shop just put in a new bottom bracket on my old fuji and didn’t use anything and it is ok, so how can you tell if you need anything?

    1. you’d really need to measure the hole first.

  3. What about using temperature shrinkage to fit the bearing (i.e.. putting the bearing in a fridge/freezer) and then letting it expand to the housing. Recommended? Avoid?

    1. Yes you can do that. It’s generally more common to heat the female surface


  5. What’s the best way to remove old retaining compound from a carbon bearing seat?

    So when would you choose 638 over 641? Is there any downside to the over 2x strength of 638?

    Presumably a higher strength would be less likely to allow the bearing to ‘spin’ and damage a frame as you identified in your look Zed video.

    As an aside, the commonly used Park RC-1 (which bizarrely Park acknowledge to be relabelled 680) is specified as suitable for up to 0.38mm with a similar 4000 vs 4500psi shear.


    1. acetone will get it off.

      1. acetone might be pretty aggressive to the epoxy used in the “carbon”, it might start softening and expand on the surface!

  6. What about headset bearings into a fully integrated carbon frame? Would you use nothing or just a small amount of grease?

    1. they should be internally greased, grease on the outside generally attracts dirt.

  7. Re paragraph “Dealing with Bike Manufacturer’s notorious Tolerances”

    “wiggle room to deal with undersized bottom bracket holes”
    should read
    “wiggle room to deal with OVERsized bottom bracket holes”

    or have I missed something?

    1. Could be both!

Please leave a reply to add to the discussion! NO NEED TO REGISTER!