How to Choose Keywords for an Engineering Journal Paper

Choose 3 to 6 keywords that cover the four things a reader searches on — the material or system, the method, the property measured, and the application — and never repeat words already in your title. Keywords are how people who have never heard of your paper find it, and most authors spend thirty seconds on them.

For an engineering journal paper this is the cheapest part of a manuscript to get right and one of the easiest to waste. A paper with generic keywords is competing with millions of records; the same paper with specific ones reaches the handful of people who needed it.

What keywords are actually for

Two different systems read your engineering journal keywords, and they use them differently:

• Indexing databases use the keyword field directly. When someone filters or searches within a database, your listed terms are part of what is matched.
• Search engines largely ignore a keyword field — they read your title, abstract and full text. But the discipline of choosing keywords well feeds straight into that, because you end up using those terms naturally in your writing.

So the keyword line is not a formality for the database alone. Working out what your readers type is the same exercise as working out what your abstract should say.

The four axes

Almost every useful engineering journal keyword set covers these four. Pick one or two from each:

A paper titled “Effect of fibre loading on tensile properties of jute-epoxy composites” already contains the material and the property. So the keywords should spend their slots elsewhere: natural fibre composite, hand lay-up, interfacial bonding, automotive interior panels, sustainable materials.

The mistake almost everyone makes

Repeating the title. It is the single commonest waste, and it happens because the title words are the ones already in your head.

Title: Corrosion behaviour of galvanised steel in marine environments
Keywords: corrosion, galvanised steel, marine environment — three slots, zero new information.

Better: zinc coating, salt spray testing, electrochemical impedance, coastal infrastructure, service life prediction. Now the paper is findable by five routes it was invisible on before.

Specific beats broad, every time

The instinct behind broad engineering journal keywords is understandable — you want the widest possible audience. It backfires. Ranking for “materials” is impossible; being the clearest result for “jute-epoxy interfacial bonding” is achievable, and those are the readers who will actually cite you.

How to find the terms your readers really use

1. Look at your own reference list. The papers you cited are the papers your readers are reading. What keywords did they use?
2. Search the phrase you were going to use. If the results are nothing like your paper, that is the wrong phrase.
3. Check for competing names. Many engineering concepts have two accepted names — include both where they genuinely coexist.
4. Ask what problem your reader has, not what you did. People search for the problem far more often than for the method.

A worked example, start to finish

Take a real-sounding paper and choose its keywords properly. The study: a fatigue test on friction stir welded aluminium joints, measuring how weld speed changes crack initiation life, aimed at aerospace panel joints.

Step 1 — write the title

Effect of welding speed on fatigue crack initiation in friction stir welded AA6061 joints

Step 2 — list what the title already contains

Welding speed, fatigue crack initiation, friction stir welding, AA6061. All four are indexed already. None of them should be a keyword.

Step 3 — fill the four axes with what is left

Five keywords, five new routes to the paper, zero overlap with the title. That took about four minutes.

Keywords for review articles and case reports

The four axes assume an experimental paper. Two other article types need a slightly different approach:

• Review articles — the “method” axis is your review scope, so use terms like systematic review or state of the art alongside the subject. Readers searching for a review use those words deliberately.
• Case reports — lead with the failure or the condition, not the analysis. Someone facing the same failure searches for the symptom, not the technique you used to diagnose it.

A five-minute keyword audit for a finished manuscript

If the paper is already written, run this before you submit. It is faster than choosing from scratch and it catches the two failures that matter.

1. Write your title and keywords side by side. Strike out every keyword whose words appear in the title. If more than one goes, you have slots to reclaim.
2. Check each surviving keyword against the four axes. Any axis with nothing against it is a route readers cannot take to your paper.
3. Search each keyword and look at the first ten results. If they are nothing like your paper, that term is not what your readers use.
4. Read your own abstract and note the terms you repeat. Words you used naturally three times are usually the right keywords — and if a keyword never appears in the abstract at all, ask whether it really describes the paper.
5. Count. Three to six. Not two, not nine.

Most manuscripts fail step 1 and step 3. Fixing just those two is the whole exercise.

What keywords cannot fix

It is worth being clear about the limits. Good keywords make a findable paper more findable. They do not rescue a paper nobody wants, they do not compensate for a vague title, and they will not put you in front of readers if the abstract does not hold their attention once they arrive.

Think of them as the last ten percent of discoverability, applied to a paper that has already earned the other ninety. That is still worth four minutes of your time — but it is four minutes, not a strategy.

Formatting and placement

At this engineering journal, keywords go immediately after the abstract, before the introduction, and there should be 3 to 6 of them. Two practical points:

• Use lower case except for proper nouns and standard designations (ASTM E8, Ti-6Al-4V).
• Keep singular and plural consistent; do not list both forms of the same term.
• No punctuation inside a keyword other than hyphens where the term genuinely carries one.

The full specification, along with the 300-word abstract limit, is on the Instructions to Author page. Our guide on writing an abstract covers how the two work together.

Frequently asked questions

How many keywords should a research paper have?

Between 3 and 6 at this journal, placed immediately after the abstract. Fewer than three wastes the field; more than six dilutes it, and some indexing systems simply truncate the list.

Should keywords repeat words from the title?

No. Search systems index the title already. Every keyword that repeats a title word is a slot you have thrown away.

Can a keyword be more than one word?

Yes, and it usually should be. Multi-word phrases such as fatigue crack propagation match how people actually search far better than a single generic word.

Should I use abbreviations as keywords?

Only alongside the expanded form, never instead of it. Give both if the abbreviation is genuinely standard in your field, because different readers search for different forms.

Do keywords affect how my paper ranks in Google?

Indirectly. The keyword field is used by indexing databases; Google reads your title, abstract and full text. Choosing terms your readers actually use helps in both places, because you naturally repeat them in the text.

Where do keywords go in the manuscript?

Immediately after the abstract, before the introduction. This journal asks for 3 to 6, separated by commas.

Related Guides

Where keywords fit with the rest of the front matter:

• How to Write an Abstract for an Engineering Research Paper
• What is an Indexed Engineering Journal? Databases Explained
• What is a DOI and Why Every Engineering Journal Paper Needs One